Mechanisms of increased vascular stiffness down the aortic tree in aging, premenopausal female monkeys

Abstract

Protection against increased vascular stiffness in young women is lost after menopause. However, little is known about vascular stiffness in older, premenopausal females, because most of the prior work has been conducted in rodents, which live for only 1–3 yr and do not go through menopause. The goal of the current investigation was to quantitate differences in stiffness down the aortic tree and the mechanisms mediating those differences in older, premenopausal (24 ± 0.7 yr) versus young adult (7 ± 0.7 yr) female nonhuman primates. Aortic stiffness (β), calculated from direct and continuous measurements of aortic diameter and pressure in chronically instrumented, conscious macaque monkeys, increased 2.5-fold in the thoracic aorta and fivefold in the abdominal aorta in old premenopausal monkeys. The aortic histological mechanisms mediating increased vascular stiffness, i.e., collagen/elastin ratio, elastin, and collagen disarray, and the number of breaks in elastin and collagen fibers were greater in the old premenopausal versus young monkeys and greater in the abdominal versus the thoracic aorta and greatest in the iliac artery. In addition, more immature and less cross-linked fibers of collagen were found in the aortas of young females. Aortic stiffness increased in old premenopausal female monkeys, more so in the abdominal aorta than in the thoracic aorta. Histological mechanisms mediating the increased aortic stiffness were augmented in the old premenopausal females, greater in the abdominal versus the thoracic aorta, and greatest in the iliac artery.

NEW & NOTEWORTHY This is the first study to examine vascular stiffness down the aortic tree in aging premenopausal females (24 ± 0.7 yr old), whereas prior work studied mainly rodents, which are short-lived and do not undergo menopause. Histological mechanisms mediating vascular stiffness in older premenopausal females increased progressively down the aortic tree, with greater increases in the abdominal aorta compared with the thoracic aorta and with the greatest increases and differences observed in the iliac artery.

INTRODUCTION

Vascular stiffness increases with aging, mediated by mechanisms involving the extracellular matrix (5, 31, 53). It is generally thought that young adult females are protected against increased vascular stiffness (37, 38, 45, 49), whereas after menopause this protection is lost (23, 38, 45, 49). However, almost nothing is known about vascular stiffness and its mechanisms in aging, pre-menopausal, females, the goal of this investigation. The absence of data on this topic is because most of the prior work in this field has been conducted in rodents (18, 28, 32, 50), which have the limitation of reaching only 1–3 yr of age and do not go through menopause (8, 51), making it difficult to translate the data to aging women. There are also human studies (4, 7, 15, 44, 52), which are limited by confounding disease states associated with aging, e.g., hypertension, atherosclerosis, and diabetes, which all have effects on aortic stiffness, independent of aging.

To address these limitations, the current investigation was conducted in nonhuman primates, which are closer to humans phylogenetically, undergo menopause, and have a lifespan of >30 yr (12, 47). Accordingly, the first goal of the current investigation was to quantitate the differences in aortic stiffness in the thoracic and abdominal aorta in young adult female nonhuman primates (7 ± 0.7 yr old) and older premenopausal ones (24 ± 0.7 yr old). It is also important to note that the studies on aortic stiffness were conducted in chronically instrumented, conscious monkeys, since anesthesia and recent surgery are known to affect vascular function (1, 17).

There are even fewer studies on the mechanisms mediating the increased vascular stiffness down the aortic tree from the thoracic to the abdominal aorta to the iliac arteries in old, premenopausal females. Accordingly, the second goal was to quantitate the differences in the histological mechanisms mediating increased vascular stiffness down the aortic tree, from the thoracic aorta, to the abdominal aorta, and to the iliac artery. Whereas most studies on vascular stiffness only examined changes in collagen and elastin in the vessels (4, 7, 15, 18, 28, 32, 34, 44, 50, 52), the current investigation also examined collagen and elastin disarray, fiber maturity, disorientation, and breaks and how these changes correlate with aortic stiffness.

MATERIALS AND METHODS

Animals

Young female (7 ± 0.7 yr old) and old premenopausal female (24 ± 0.7 yr old) Macaca fascicularis and Rhesus macaque monkeys were studied. The old premenopausal female monkeys were premenopausal (regular monthly menstrual cycles), which is consistent with prior reports in these species (2, 31, 47). Monkeys used were from the Oregon National Primate Research Center, the University of Wisconsin Primate Center, and Alpha Genesis, Inc. The animals used in the current study were maintained in accordance with the Guide for the Care and Use of Laboratory Animals (National Research Council, 8th ed., 2011). The data from Fascicularis and Rhesus were similar. Experimental protocols were approved by the Institutional Animal Care and Use Committee at Rutgers, New Jersey Medical School; Oregon National Primate Research Center, the University of Wisconsin Primate Center; and Alpha Genesis, Inc.

Aortic Stiffness

Monkeys were tranquilized with ketamine hydrochloride (2 to 3 mg/kg im), anesthetized with thiamylal sodium (5 to 10 mg/kg iv), and maintained with isoflurane (0.5 to 1.5 mL/100 mL in oxygen) during surgery. Aortic stiffness was compared in the thoracic and abdominal aorta; it was not feasible to measure stiffness in the iliac artery, which was too small to implant piezoelectric crystals for measuring iliac diameter. With a left thoracotomy approach at the level of the fourth intercostal space, Tygon catheters were implanted in the descending thoracic aorta, and one pair of piezoelectric crystals was implanted on opposing surfaces of the proximal third of the descending thoracic aorta at the level of the sixth intercostal space. With a midline laparotomy approach, another pair of piezoelectric crystals was implanted on opposing surfaces of the abdominal aorta 1 to 2 cm below the renal arteries. Measurements were made 3 days later in the conscious state, using a tether system, after full recovery from surgery, as certified by the attending veterinarian and evidence of normal activity and food and water intake. Hemodynamic data in these monkeys are similar to those in monkeys that were studied 2 to 3 wk after surgery (3, 35, 39). The tether system provided access to aortic pressure and diameter and enabled the investigators to perform the procedures from a separate room, not observed by the monkeys, and without the need for further restraint of the animals. Data were recorded on LabChart 7 software (AD Instruments) with a Powerlab/8SP (AD Instruments) data acquisition system. Aortic pressure was measured with the implanted aortic catheter and strain gauge manometer. Diameters were measured with an ultrasonic dimension gauge and with piezoelectric ultrasonic crystals implanted on opposing surfaces of the thoracic and abdominal aorta. Aortic stiffness (β) was computed as ln(aortic systolic pressure/aortic diastolic pressure)/aortic strain (53). Aortic strain was calculated as [systolic aortic diameter (D_s) − diastolic aortic diameter (D_d)]/D_d in the thoracic and abdominal aorta, respectively.

Histology

Samples for histology were collected from the proximal third of the descending thoracic aorta (9 each in old premenopausal and young monkeys) and from the abdominal aorta (9 each in old premenopausal and young monkeys) and from the left common iliac artery (6 each in old premenopausal and young monkeys). Samples were preserved in 10% buffered formalin and were embedded in paraffin in a cross-sectional orientation; 5-μm-thick aortic cross-sectional slices were cut and mounted onto glass slides. Histological analyses were performed blindly by two different investigators.

Collagen and elastin density.

Picrosirus red staining (Sigma-Aldrich Lot: 197378-100G) was used to visualize medial collagen fibers using both circular, cross-polarized light and visible light microscopy. Aldehyde fuchsin staining (basic fuchsin, 70% ethyl alcohol, and hydrochloric acid paraldehyde) was used to visualize elastin fibers within the aortic wall. Collagen and elastin density were determined within each sampled field (15 fields/ring) by direct pixel quantification (MATLAB software).

Collagen and elastin disarray, breaks, and orientation.

The disarray of elastin and collagen in the aorta and iliac artery media was determined by loss of circumferential fiber orientation and increased dispersion/randomness of fibers and was graded from a level of 1–10 among all the aortic and iliac rings of both young and old pre-menopausal monkeys under ×10 magnification, as previously described (53). Level 1 reflected almost all collagen and elastin fibers evenly distributed and aligned circumferentially. Level 10 reflected samples with 95–100% of elastin or collagen fibers disrupted, distorted, or aggregated. Grading of samples was performed blinded by two different investigators.

Fiber breaks and orientation were also examined separately. Images were taken under ×20 magnification for counting fiber breaks and orientation analysis. The number of breaks in collagen and elastin were counted manually, and the data are presented as the total number of breaks divided by total number of fibers. For fiber orientation, the images were cropped such that only the arterial media was visible and that the circumferential direction of the arterial media roughly corresponded to the horizontal plane of the image. With MATLAB Image Processing Toolbox, color thresholding functions were defined and used to transform the cropped color images of the slides to black and white images, where the white regions corresponded to fibers and black corresponded to background. In a custom MATLAB (Mathworks R2019B) script, the black and white images were skeletonized as described by Koch et al. (20) to produce a binary representation of the fibers. At this point, each contiguous region of white pixels represented an individual fiber fragment. The orientation of each contiguous region was obtained, and the fiber angles were calculated using the MATLAB region props function. Histograms of the individual fiber angles were plotted to obtain a frequency distribution of fiber angle. Color thresholding in MATLAB software was also used in conjunction with circular, cross-polarized light microscopy for measurements of green and yellow-red birefringence in collagen fibers, which reflected the maturity of the fibers (31, 46, 53).

Data Analysis and Statistics

All results except fiber disorientation are expressed as means ± SE; comparisons between three individual groups considering age and region and the interactions between them were analyzed using a two-way ANOVA test supplemented by a Bonferroni test to reduce type 1 errors with the ANOVA. Fiber disorientation data are given as the standard deviation of the fiber orientation angles for each group ± SE. Comparisons between fiber disorientation groups were made using the F test for variance. For linear regressions, the mean values and standard error from each group were used for the Pearson correlation analysis.

RESULTS

Aortic Stiffness

An illustration of the instrumentation used to measure aortic stiffness in chronically instrumented, conscious, young and old female monkeys is shown in Fig. 1. The aortic pulse pressure was increased in the old premenopausal monkeys (48 ± 2.7 mmHg) compared with young monkeys (33 ± 2.5 mmHg). There was no difference in aortic systolic pressure between young (104 ± 5 mmHg) and old premenopausal (106 ± 4 mmHg) female monkeys (vs.), but heart rate was slightly higher in old premenopausal (172 ± 6 beats/min) versus young (157 ± 9 beats/min) female monkeys. The calculated aortic stiffness index (β) was greater in the abdominal aorta than in the thoracic aorta and also greater in old premenopausal versus young monkeys in both the thoracic aorta (20 ± 1.8 vs. 8 ± 1.1) and abdominal aorta (39 ± 7.5 vs. 13 ± 1.3).

Fig. 1.

An aortic pressure catheter and ultrasonic diameter transducers were implanted to measure aortic stiffness in the thoracic and abdominal aorta, as shown in the diagram on left, enabling continuous phasic measurements of aortic pressure and diameter in young adult (n = 10 for thoracic aorta, n = 3 for abdominal aorta) and old premenopausal female monkeys (n = 11 for thoracic aorta, n = 4 for abdominal aorta), as shown on right. Note the greater aortic pulse pressure but reduced pulse diameter changes in the old premenopausal monkeys, reflecting their increased aortic stiffness.

Elastin and collagen density

Elastin density was decreased in the thoracic aortic media of old premenopausal female monkeys (23 ± 1.5%) versus young females (30 ± 1.0% P < 0.05) and progressively decreased in old premenopausal females down the aortic tree from the thoracic (23 ± 1.5%) to the abdominal aorta (18 ± 1.5%) to the iliac artery (14 ± 1.0%) (Fig. 2A and Table 1). Collagen density was increased in the thoracic aorta of old premenopausal females (24 ± 1.6%) versus young females (15 ± 1.7%, P < 0.05), and progressively greater increases were observed down the aortic tree in the old premenopausal females from the thoracic (24 ± 1.6%) to the abdominal aorta (31 ± 1.4%) to the iliac artery (44 ± 1.3%) (Fig. 2B and Table 1). The collagen/elastin ratio was increased more in the thoracic aorta of old premenopausal females (1.0 ± 0.1) versus young females (0.5 ± 0.1, P < 0.05), and progressively greater increases were observed down the aortic tree in old premenopausal females from the thoracic (1.0 ± 0.1) to the abdominal aorta (1.8 ± 0.2) to the iliac artery (3.2 ± 0.2) (Fig. 2C and Table 1). Changes in collagen/elastin ratio were linearly correlated with changes in aortic stiffness in the thoracic and abdominal aorta (Fig. 2D). Low-power images showing the reduced elastin density and increased collagen density in the old premenopausal females are shown in Fig. 3.

Fig. 2.

A–C: elastin density decreased (A) and collagen density (B) and the collagen/elastin ratio (C) increased progressively down the aortic tree from the thoracic to the abdominal aorta to the iliac artery and were greater in old premenopausal female monkeys (solid green bars) vs. young female monkeys (open bars). D: the increases in the collagen/elastin ratio correlated linearly with aortic stiffness in the thoracic and abdominal aorta. For old premenopausal females, n = 12 for elastin in both the thoracic and the abdominal aorta, and n = 14 for collagen in both the thoracic and abdominal aorta. For young females, n = 9 for both elastin and collagen in both the thoracic and abdominal aorta. In the iliac artery (n = 6 for both young and old premenopausal monkeys). *P < 0.05 by the 2-way ANOVA test, supplemented by a Bonferroni test to reduce type 1 errors with the ANOVA. A Pearson correlation test was also used.

Table 1.

Histological changes were greater in the AA than in the TA and greatest in the iliac artery in old premenopausal female monkeys

Histological Parameter	%Changes in AA Greater Than in TA	%Changes in Iliac Artery Greater Than in AA
Elastin density	18 ± 5	26 ± 5
Collagen density	32 ± 6	43 ± 4
Collagen/elastin ratio	67 ± 15	81 ± 12
Elastin disarray	60 ± 15	44 ± 7
Collagen disarray	31 ± 8	37 ± 4
Elastin No. of breaks	67 ± 11	140 ± 23
Collagen No. of breaks	30 ± 10	111 ± 9
Fiber disorientation elastin	48 ± 14	12 ± 15
Fiber disorientation collagen	9 ± 7	4 ± 26

Values are means ± SE. AA, abdominal aorta; TA, thoracic aorta.

Fig. 3.

Comparison of elastin (A and B) and collagen (C and D) fiber density in the young female and the old premenopausal female abdominal aorta. In old premenopausal monkeys, there were decreases in elastin density and increases in collagen density.

Elastin and Collagen Disarray, Fiber Disorientation, and Number of Breaks

We defined the disarray of elastin and collagen fibers as the loss of circumferential orientation and the increased dispersion/randomness/breaks of fibers. Both elastin disarray and collagen disarray progressively increased down the aortic tree and were roughly twice as great in old premenopausal females versus young females in the thoracic and abdominal aorta and in the iliac artery (P < 0.05; Fig. 4, A and B, and Table 1). To illustrate these differences, high-power histological examples of the changes down the aortic tree are shown for elastin density, breaks and disarray, and collagen density (breaks and disarray in Figs. 5 and 6). Changes in elastin disarray and collagen disarray showed linear correlations, with changes in aortic stiffness in the thoracic and abdominal aorta (Fig. 4, C and D).

Fig. 4.

A and B: both elastin (A) and collagen (B) disarray increased progressively down the aortic tree from the thoracic to the abdominal aorta to the iliac artery and were greater in old premenopausal female monkeys (solid green bars) than in young female monkeys (open bars). C and D: these changes correlated linearly with measurements of aortic stiffness in the thoracic and abdominal aorta; n = 16 for old premenopausal females in thoracic aorta, n = 14 for old premenopausal females in the abdominal aorta, and n = 9 for young monkeys in both the thoracic and abdominal aorta, whereas in the iliac artery, n = 6 for both young and old premenopausal monkeys. *P < 0.05 by the 2-way ANOVA test, supplemented by a Bonferroni test to reduce type 1 errors with the ANOVA. A Pearson correlation test was also used.

Fig. 5.

Examples of elastin fiber disarray, breaks, and fiber disorientation in the thoracic and abdominal aorta and the iliac artery of young females (A, C, and E) as compared with those of old premenopausal females (B, D, and F). In old premenopausal monkeys, elastin disarray, fiber breaks, and fiber disorientation increased whereas elastin density decreased down the aortic tree from the thoracic aorta to the abdominal aorta to the iliac artery more than in the young females. Green arrows indicate the change of fiber angle from the starting point of a fiber toward the end point of a fiber, and black arrows indicate the fiber breaks.

Fig. 6.

Examples of collagen fiber disarray, breaks, and fiber disorientation in the thoracic and abdominal aorta and the iliac artery of young females (A, C, and E) as compared with those of old premenopausal females (B, D, and F). In old premenopausal monkeys, collagen disarray and fiber disorientation increased whereas collagen density decreased down the aortic tree from the thoracic aorta to the abdominal aorta to the iliac artery more than in the young females. Green arrows indicate the change of fiber angle from the starting point of a fiber toward the end point of a fiber, and the black arrows indicate the fiber breaks.

Collagen fiber composition was also examined by circular cross-polarized light microscopy. In young female monkeys, there was an increase in green birefringence, which reflected an abundance of immature or thin collagen fibers, compared with old premenopausal female monkeys (Fig. 7A). Old premenopausal female monkeys exhibited increased yellow-red birefringence, reflecting thicker, more extensively cross-linked, mature collagen fibers as compared with young monkeys in the thoracic and abdominal aorta and the iliac artery (Fig. 7B). Images of these differences in collagen fibers between young and old premenopausal female monkey aortas are shown in Fig. 7, C and D.

Fig. 7.

A and B: the higher mean ratio of thin green fibers to total fibers, which indicates newly synthesized collagen in the thoracic, abdominal, and iliac artery of young monkeys (open bars) compared with old premenopausal monkeys (solid green bars) (A), whereas the higher ratio of thick, yellow-red fibers (B) indicates more cross-linked, mature collagen in the thoracic, abdominal, and iliac artery of old premenopausal monkeys compared with young monkeys. For the thoracic aorta: young females, n = 7; old females, n = 8; for abdominal aorta and iliac artery, n = 5/group. C and D: illustrative images from young and old premenopausal female monkey aortas demonstrating these differences are shown. White arrows indicate the yellow-red collagen fibers. *P < 0.05 by the 2-way ANOVA test, supplemented by a Bonferroni test to reduce type 1 errors with the ANOVA. A Pearson correlation test was also used.

In addition, the number of breaks in elastin fibers and n collagen fibers increased progressively down the aortic tree and were roughly twice as great in old premenopausal females versus young females (P < 0.05; Fig. 8, A–D, and Table 1).

Fig. 8.

A and B: the number of breaks in elastin fibers (A) and in collagen fibers (B) increased progressively down the aortic tree from the thoracic to the abdominal aorta to the iliac artery and was greater in old premenopausal female monkeys (solid green bars) than in young female monkeys (open bars). C and D: these changes correlated linearly with measurements of aortic stiffness in the thoracic and abdominal aorta. For thoracic and abdominal aorta, n = 9/group; for iliac artery, n = 6/group. *P < 0.05 by the 2-way ANOVA test, supplemented by a Bonferroni test to reduce type 1 errors with the ANOVA. A Pearson correlation test was also used.

Elastin and collagen disorientation, as reflected by the standard deviation of fiber angles, was increased more in the old premenopausal females than in the young females and also increased progressively down the aortic tree, with the greatest differences in the iliac artery, where the standard deviation of elastin fiber angles was 47.2 ± 1.5° in the old premenopausal females and 31.2 ± 1.3° in the young females and the standard deviation of collagen fiber angles was 43.1 ± 1.7° in the old pre-menopausal females and 31.8 ± 1.4° in the young females (Fig. 9, A–D, Table 1).

Fig. 9.

A and B: standard deviation (SD) of the fiber angle indicating fiber disorientation of elastin (A) and collagen fibers (B) increased progressively down the aortic tree from the thoracic to the abdominal aorta to the iliac artery and were greater in the old premenopausal monkeys (solid green bars) than in the young female monkeys (open bars). C and D: these changes correlated linearly with measurements of aortic stiffness in the thoracic and abdominal aorta. There are 612 ± 82 fibers in young females of each region and 625 ± 91 fibers in old premenopausal female of each region. *P < 0.05 by the 2-way ANOVA test, supplemented by a Bonferroni test to reduce type 1 errors with the ANOVA. A Pearson correlation test was also used. Histological examples of collagen and elastin density and disarray in the old premenopausal vs. young females are shown in Figs. 5 and 6.

Additionally, frequency distribution histograms of fiber angles for collagen and elastin fibers in the thoracic aorta (Figs. 10, A and B), abdominal aorta (Fig. 10, C and D), and iliac artery (Figs. 10, E and F) were compared between young and old premenopausal female monkeys. The central bin of the fiber angle histogram (0°) indicates no fiber disorientation, with increasing fiber disorientation progressing to the left and to the right sides away from the central bin. Old premenopausal female monkeys exhibited a wider distribution of fiber angles in the aorta and especially in the iliac artery, reflecting greater fiber disorientation.

Fig. 10.

Frequency distribution histograms of collagen (A, C, and E) and elastin (B, D, and F) fiber angles, reflecting orientation in the thoracic aorta (A and B), the abdominal aorta (C and D), and the iliac artery (E and F), are compared in young (open bars) and old premenopausal (solid bars) female monkeys. The wider distribution of fiber angles in old premenopausal female monkeys, reflected by the solid bars compared with young female monkeys (open bars), indicates that fiber disorientation was progressively greater down the aortic tree in the old premenopausal female monkeys. In young females for both elastin and collagen, n = 6, 5, and 4 for thoracic aorta, abdominal aorta, and iliac artery, respectively. In old premenopausal females, n = 5, 5, and 4 for both collagen for elastin in the thoracic aorta, abdominal aorta, and the iliac artery, respectively. There are 612 ± 82 fibers in young females of each region and 625 ± 91 fibers in old premenopausal females of each region.

DISCUSSION

Increases in arterial stiffness lead to a rise in cardiac afterload, which in turn results in cardiac decompensation, and are also involved in the pathogenesis of hypertension (25). Although it is generally thought that vascular stiffness is less pronounced in premenopausal females than males of the same age and that the protection in females is lost after menopause (23, 38, 45, 49), it should be noted that the published data on this topic are controversial, with many studies failing to note that male vascular stiffness is higher than female vascular stiffness (4, 11, 13, 14, 37). Other studies showed no difference between age-matched men and women at all ages with stiffness measured by pulse wave velocity (6, 23, 26, 37, 43) or by MRI (14). Other studies have shown that age affects vascular stiffness in women more severely than in men, with women having a steeper increase in aortic stiffness than age-matched males measured by pulse wave velocity (22, 38) or by augmentation index (6, 16, 26). In addition, there are studies in humans indicating that estrogen replacement in old women results in a lower stiffness index (27, 36), but no study has explored how estrogen contributes to arterial stiffness down the aortic tree. Our prior study found greater increases in aortic stiffness in the abdominal aorta than the in thoracic aorta in old male monkeys, with increases in stiffness greater than observed in young male monkeys (53).

Almost all prior studies on aortic stiffness measured stiffness at only one point in the aorta, whereas in the current investigation aortic stiffness was measured in chronically instrumented nonhuman premenopausal primates in both the thoracic and the abdominal aorta, revealing that aortic stiffness was higher in the abdominal aorta compared with the thoracic aorta and in both cases significantly higher than in young female monkeys. This is the first study to find these differences in aortic stiffness in premenopausal female animals. In two other studies, pulse wave velocity was higher in the thoracic aorta of old premenopausal females compared with young females (11, 40), whereas in another study it was not (23).

It is important to note that not all studies have results consistent with ours. For example, although the majority of the literature indicates that stiffness is greater in the abdominal aorta compared with the thoracic aorta, one study conducted in humans found the reverse with aging (33). Therefore, it was important to resolve the controversy in the best animal model for aging in human females, the nonhuman primate. This model is superior to rodents, where most of the current work has been done, since rodents only live 1–3 yr and the females do not undergo menopause, whereas macaques used in this study live to 30–40 yr and undergo menopause similar to that observed in human females. The female monkeys in this study were 24 yr old, still having regular monthly menstrual cycles, i.e., well below the age of menopause in Macaques (2, 31, 47).

The most novel aspect of this investigation related to finding progressively increased differences in histological mechanisms responsible for increased vascular stiffness between the old premenopausal and young female monkeys down the aortic tree, with greater differences in the abdominal aorta than in the thoracic aorta and even greater differences in the iliac artery than in the abdominal aorta (Table 1). As with aortic stiffness, almost all prior studies have examined these histological mechanisms at only one location in the aorta. More importantly, this is the first investigation comparing changes in a regional artery, i.e., the iliac artery, with changes in the aorta in females.

Although iliac artery stiffness was not measured directly in the current study, it is likely that it was increased even more than the stiffness in the abdominal aorta, based on the linear correlations between aortic stiffness and histological mechanisms mediating aortic stiffness, which were greater in the abdominal versus the thoracic aorta. Because the histological mechanisms were greater in the iliac artery than in the abdominal aorta, it is likely that iliac artery stiffness was also greater than abdominal aortic stiffness.

The most commonly studied histological mechanism mediating vascular stiffness involves increases in collagen content and decreases in elastin content of the arterial media. We found that the collagen/elastin ratio nearly doubled in the thoracic aorta and tripled in the abdominal aorta and iliac artery in old premenopausal female monkeys when compared with young female monkeys. There are no other studies on this topic demonstrating greater increases in collagen and decreases in elastin down the aortic tree from the thoracic to the abdominal aorta to the iliac artery comparing old premenopausal and young female monkeys or any other young and old animal models. Furthermore, collagen fiber composition was examined in the aortas, with more immature and less cross-linked collagen fibers exhibited in young female monkeys than in old premenopausal female monkeys. There have been other studies measuring collagen and elastin down the aortic tree (4, 9, 19, 53), but in most of these studies the measurements were not to the level of the iliac arteries (4, 19, 53). In one other study (9), the femoral artery was also examined, but not in females or in an aging model. Four studies have found that the increases in collagen and decreases in elastin were greater in the abdominal aorta than in the thoracic aorta (4, 9, 19, 53), but these studies were either in males (9, 53) or mixed sexes (4, 19) and, therefore, cannot be compared with the current investigation solely in aging premenopausal females.

Extracellular matrix disarray is another important but less well-studied mechanism responsible for increased aortic stiffness (53). Characteristics of extracellular matrix disarray are disruption of circumferential orientation and breakage of the collagen and elastin fibers (53). Examples are shown in Figs. 5 and 6. The collagen and elastin fiber disarray were progressively greater down the aortic tree from the thoracic to the abdominal aortas to the iliac artery and twice as great in old premenopausal females compared with young female monkeys. Moreover, the increased disarray was significantly and linearly correlated with increased stiffness, which was measured in the thoracic and abdominal aorta. The number of breaks in collagen and elastin fibers and the degree of their fiber disorientation, as reflected by the standard deviation of fiber angles, are key components of extracellular matrix disarray and contribute to increased aortic stiffness with aging (41). We found that the number of breaks and the fiber disorientation in both elastin and collagen fibers were greater in old premenopausal female monkeys than in young female monkeys, and these differences were progressively greater down the aortic tree, with the greatest changes in aging occurring in the iliac artery. Histogram analysis of both elastin and collagen fiber angles suggested a wider fiber angle distribution in old premenopausal female monkeys, especially in the iliac artery. This is the first study to analyze collagen and elastin disarray, fiber disorientation, and the number of breaks in collagen and elastin fibers down the aortic tree from the thoracic to abdominal aorta to iliac artery. Furthermore, prior studies of vascular extracellular matrix disarray were conducted either only in males (53) or in groups with mixed sexes (10, 21, 24, 29), not focusing on females, and focused exclusively on a single location of the aortic tree. Although some studies investigated the histogram of fiber angles from the aorta, they were all in disease states and did not compare different locations in the aging aorta (20, 30, 42). In addition, our previous study assessed aortic type IV collagenase, matrix metalloproteinase-2 (MMP-2), and found increased MMP-2, MT1-MMP, and angiotensin II activity in old monkeys, with no change in tissue inhibitor of MMP (TIMP-2) antibody-staining fraction, suggesting that the age-associated arterial remodeling results in MMP-2 activation and increased angiotensin II signaling (48).

In conclusion, the considerable controversy related to changes in aortic stiffness with aging, sex differences, and differences down the aortic tree was resolved by studying this problem in old premenopausal and young nonhuman primates without associated vascular disease, with the most direct measurements of aortic dimensions and pressure available, and conducted in conscious animals without the confounding effects of anesthesia and recent surgery. Aortic stiffness increased in old premenopausal female monkeys compared with young female monkeys more in the abdominal aorta than in the thoracic aorta. Furthermore, the mechanisms mediating these changes not only involved decreases in elastin density and increases in collagen density but also increased elastin and collagen fiber disarray and breaks as well as increased collagen fiber maturity, all contributing to the increased vascular stiffness in aging female monkeys. All of these mechanisms mediating vascular stiffness were more pronounced in the abdominal versus the thoracic aorta and were greatest in the iliac artery, as summarized in Table 1. These data show that major increases in vascular stiffness and the mechanisms mediating these increases occur in aging female nonhuman primates before menopause.

GRANTS

This study was supported by National Institutes of Health Grants R21-AG-053514, R01-HL-119464, R01-HL-137405, R01-HL-102472, R01-HL-106511, R01-HL-130848, R01-HL-137368, and UL1-TR-003017.

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

D.E.V. and S.F.V. conceived and designed research; D.B., R.K.K., T.M., and T.B. performed experiments; D.B., J.Z., and T.B. analyzed data; D.B., J.Z., M.O., T.B., S.P.B., and D.E.V. interpreted results of experiments; D.B., J.Z., and T.B. prepared figures; D.B., J.Z., and T.B. drafted manuscript; D.B., R.K.K., J.Z., M.O., T.B., K.N., S.P.B., D.E.V., and S.F.V. edited and revised manuscript; D.B., D.E.V., and S.F.V. approved final version of manuscript.