Effect of ovariectomy and Sideritis euboea extract administration on large artery mechanics,morphology,and structure in middle-aged rats

Abstract

BACKGROUND:

Arterial function is regulated by estrogen, but no consistent pattern of arterial mechanical remodeling in response to depleted estrogen levels is available.

OBJECTIVE:

To examine long-term effects of ovariectomy (OVX) on the mechanical properties, morphology, and histological structure of the carotid artery in middle-aged rats and a potentially protective effect of Sideritis euboea extract (SID), commonly consumed as “mountain tea”.

METHODS:

10-month-old female Wistar rats were allocated into control (sham-operated), OVX, $OVX + SID$ , and $OVX + MALT$ (maltodextrin; excipient used for dilution of SID) groups. They were sacrificed after 6 months and their carotid arteries were submitted to inflation/extension tests and to dimensional and histological evaluation.

RESULTS:

Remodeling in OVX rats was characterized by a decreased in situ axial extension ratio, along with increased opening angle, thickness, and area of the vessel wall and of its medial layer, but unchanged lumen diameter. Compositional changes involved increased elastin/collagen densities. Characterization by the “four-fiber” microstructure-motivated model revealed similar in situ biaxial response of carotid arteries in OVX and control rats.

CONCLUSIONS:

Carotid artery remodeling in OVX rats was largely consistent with hypertensive remodeling, despite the minor arterial pressure changes found, and was not altered by administration of SID, despite previous evidence of its osteo-protective effect.

Keywords

Carotid artery mechanical testing material characterization opening angle collagen

1. Introduction

The lower incidence of cardiovascular disease in premenopausal women as compared to age-matched men and the sharp rise after menopause, as well as the beneficial effects of estrogen therapy in postmenopausal women indicate a protective role of estrogen on the cardiovascular system [1]. Several mechanisms have been put forth as the basis for these protective effects. Estrogen alters serum lipid concentrations in favor of high-density lipoproteins (systemic actions), while, as part of its direct actions on blood vessels, enables vasodilation (short-term) and inhibits the blood vessel response to injury and the development of atherosclerosis (long-term effects) [2–4].

Surprisingly, although it is generally thought that depleted estrogen levels in postmenopausal women are associated with greater age-related arterial stiffening compared to men and that increased arterial stiffness is an independent risk factor for cardiovascular disease, only few clinical studies have addressed the mechanical remodeling of arteries after menopause [5–7]. As regards the experimental (animal) studies, changes in arterial wall properties have been documented in ovariectomized rodents, serving as common models of estrogen deficiency, but there is clearly a lack of unanimity vis-à-vis the exact nature of the changes in wall composition, dimensions, and mechanical properties [8–14]. Even more importantly, the young age of the ovariectomized animals in all of the previous studies does not reflect the age of women at menopause and the possibility of distinct modes of arterial remodeling or even of impaired remodeling capability in middle-aged animals cannot be dismissed, given the known changes with aging in arterial phenotype and hemodynamics, i.e. the blood pressure and flow conditions that also regulate arterial remodeling.

Accordingly, it was the primary objective of the experiments described herein to address prior inconsistencies and to define changes in the mechanical properties of the carotid arteries associated with surgical ovariectomy in the middle-aged female rat. Our attention was focused on the carotid arteries, because their mechanical properties, along with those of the other central arteries, modulate cardiovascular homeostasis through the buffering of the pulsatile output of the heart and its translation into a steady flow in the periphery, constituting a prognostic factor in cardiovascular disease [15]. Our secondary objective was to examine the effect of Sideritis euboea (SID; Mediterranean mountain tea) extract administration to determine whether it produces a vasculo- and/or cardio-protective effect. Previous publications have demonstrated that the extract from this plant has bone mineralizing properties in vitro [16], as well as bone density and strength protective effects in vivo in the mature ovariectomized rat model that closely mimics postmenopausal osteoporosis in women [17]. This and several other plant extracts appear to be potential alternative options to pharmaceutical prevention of postmenopausal osteoporosis in women [18,19]. Third, detailed morphological and histological studies were conducted to measure the arterial geometrical and compositional changes, and to correlate them with the changes in the mechanical properties, issues that have not been investigated in this animal model to date.

2. Materials and methods

2.1. Animal groups, induction of ovariectomy, and administration of extracts

The study protocol was approved and licensed (permit no. K 5181/2006) by the General Veterinary Services and the animals’ care and use was according to Greek legislation (Presidential Decree 160/1991) in compliance to the Directive 86/609/EEC. Thirty-two mature (10-month-old) healthy female Wistar rats were obtained from the registered breeding facility of the Biomedical Research Foundation of the Academy of Athens (Athens, Greece). The rats were housed four to a cage under standard laboratory conditions, i.e. 19–22°C temperature, 55–65% relative humidity, 15 air changes/hour, and a light/dark cycle of 12/12 hours. The rats were allocated to four groups: control ( $n = 8$ ), ovariectomy (OVX, $n = 8$ ), OVX plus maltodextrin ( $OVX + MALT$ , $n = 8$ ), and OVX plus maltodextrin plus SID ( $OVX + SID$ , $n = 8$ ). Maltodextrin was the excipient used for the dilution of SID. Its unknown and potential effect on the vascular wall made it necessary to include its administration to a separate animal group. Following an acclimatization period of 10 days after arrival, the rats in the control group were sham operated, whereas those in the remaining groups were ovariectomized. Under anesthesia with intramuscular injection of ketamine (100 mg/kg) and xylazine (5 mg/kg) and aseptic conditions, bilateral OVX was performed through a midline ventral incision. The incision of the peritoneum and skin was closed in layers by single interrupted sutures.

MALT was administered to the $OVX + MALT$ group and MALT with SID diluted into it was administered to the $OVX + SID$ group, both available ad libitum in the rats’ drinking water immediately after OVX. The bottles were shaken twice daily to avoid sedimentation of the extract. The preparation of the extract was made twice weekly, adjusting the concentration according to the last recorded consumption. Extract and water consumption were monitored twice a week in all rat cages. The mean daily fluid intake was 18 ml per $OVX + MALT$ and $OVX + SID$ rats. The mean daily extract intake of the $OVX + SID$ rats was 330 mg/kg of body weight. Body weight and food consumption were measured on a weekly basis. The administration period lasted for 6 months, during which the rats were observed closely. Blood pressure and heart rate were measured in the unanesthetized rats by an indirect tail-cuff method (non-invasive blood pressure system NIBP Panlab LE5002; Panlab/Harvard Apparatus, Barcelona, Spain). Six readings were taken and averaged at each measurement.

2.2. Euthanasia and tissue collection

At the end of the 6 month postoperative administration period and following euthanasia by an overdose of anesthesia, an autopsy was conducted to inspect the rats for malignancies or other pathological findings, to substantiate the success of OVX, and for the resection of uteri and common carotid arteries. The success of OVX was substantiated by failure to detect ovarian tissue and by observing notable atrophy of the uterine horns. The uteri were excised and weighed. The left common carotid artery was exposed via a midline neck incision and two superficial surgical knots were tied immediately distal to the aortic arch and immediately proximal to the carotid bifurcation, taking care not to exert excessive loads. The axial in situ length of the left common carotid artery was determined by measuring the distances between the knots and loose connective tissues were cautiously trimmed. During measurements and excision, the head of the rats was held in a natural position, namely almost aligned with the body. The carotid artery including the knots was then excised, put in a cuvette with Krebs–Ringer solution at 37°C to stabilize for $\sim 1 h$ , and its ex situ unloaded length was measured as the distance between the knots outside of the body. The axial in situ extension ratio, defined as the ratio of in situ to ex situ segment lengths, was computed. Carotid ring specimens of 3-mm width from the proximal edge of the excised vessel were fixed with formaldehyde. The left-over carotid specimen was kept for mechanical studies within 5 h after euthanasia, during which it was stored in Krebs–Ringer solution at 4°C and slowly warmed up to 37°C before testing.

2.3. Histology

Paraffin-embedded transverse sections of the carotid artery were stained with hematoxylin-eosin, Sirius red, and orcein, using standard techniques and reagents. All measurements were carried out using image-analysis software (Image-Pro Plus v4.5; Media Cybernetics Inc, Silver Spring, MD, USA) on images captured by a digital camera (Altra20; Soft Imaging System, Munster, Germany) connected to a light microscope (Olympus CX31; Olympus, Tokyo, Japan), as has been previously detailed [20–22]. Vessel dimensions were measured on hematoxylin-eosin stained sections under suitable magnification, specifically $\times 10$ for inner and outer diameter, and $\times 40$ for thickness and area measurements. Measurements of wall thickness as well as of medial and adventitial thickness were carried out at ten equidistant points around the perimeter of the carotid, and an average value calculated. Quantification of the extent of staining via color segmentation was carried out with the software again at ten equidistant areas. Specific colors of interest, i.e. red for collagen in Sirius red and orange for elastin in orcein-stained sections, were isolated and quantified as the percentage area of the total field staining positive. Cell area density was determined on assuming that ground substance occupied a negligible fraction of wall area, so that $cell density = 100 - elastin - collagen density$ . Area densities for the entire wall were calculated from those of the two layers and their areas. The latter were measured using the hand-tracing method, with the tunica media defined between the internal and external elastic lamina, and the tunica adventitia defined between the external elastic lamina and the external edge of dense adventitia. Cell number density was assessed by enumerating the nuclear profiles within ten equidistant areas around the perimeter and calculating the mean number per unit surface area. Values for each specimen were averages from three sections.

2.4. Mechanical studies

2.4.1. Geometrical measurements

Prior to examining the inflation/extension characteristics of the vessel segments, thin ( $\sim 1 mm$ ) rings were sliced from their ends for geometrical studies, as in previous studies from our group [20–22]. These were put in a Petri dish with Krebs–Ringer solution at 37°C and photographs were taken of the no-load state with a digital camera (model E400; Olympus Optical Co Ltd, Tokyo, Japan) connected to a stereomicroscope (Stemi 2000C; Carl Zeiss Optical, Chester, VA, USA). Subsequently, the rings were cut open anteriorly and left to equilibrate for 30 min before being photographed again in the zero-stress state. Measurements of ring thickness, of inner and outer perimeters in both states, and of opening angle were carried out by the Sigma Scan Pro 5 image-analysis software (Aspire Software International, Leesburg, VA, USA). Cumulative results from the proximal and distal rings were considered for processing.

Fig. 1.

Schematic diagram of the experimental apparatus used for inflation/extension testing.

2.4.2. Inflation/extension testing

The residual carotid artery segments were mounted at their in situ length in the experimental apparatus, previously described in detail [20–22] and shown schematically in Fig. 1; with the implicit assumption that the axial in situ extension ratios of the segments used for mechanical testing were identical to those of the entire carotids harvested. The segments were immersed in the temperature-controlled bath of the apparatus and horizontally mounted on two adaptors, consisting of 20-gauge needles plugged on stainless steel cannulae. One of those was kept fixed whereas the second was connected to an isometric force transducer (Fort 100; World Precision Instruments, Hertfordshire, UK), measuring the axial force exerted to the segments. This adapter was suspended from a manifold on a screw-operated slide mechanism, used to regulate the length of vessels and their position in the bath with a micrometer (Tesa Technology, Renens, Switzerland). Length measurement was performed as the separation distance between the cannulae, justified by the large aspect ratio of the arterial segments (their average no-load length, $13.9 \pm 0.5 mm$ , being over ten times greater than their average outer diameter; Table 2 in Section 3.3.1), presumably making negligible the errors caused by edge-effects. The outer diameter of the arterial segments was measured by a laser beam micrometer (LS-3100; Keyence Corp, Osaka, Japan), placed on the midsection of the segments. Lumen pressure was measured by a pressure transducer (BLPR; World Precision Instruments, Hertfordshire, UK). The bath was filled with Krebs–Ringer solution, oxygenated with 95% O₂ and 5% CO₂, and sustained at 37°C. With the addition of papaverine ( $100 μ M$ ) to the bath, passive properties were examined under negligible muscle tone. The diameter, pressure, and force signals were amplified via a four-channel transducer-amplifier (Transbridge 4M; World Precision Instruments) and constantly recorded on a data acquisition card (DAQmx 6009; National Instruments, Austin, TX, USA). All data signals were synchronized and stored in a nearby computer via an acquisition interface (Labview v7.1; National Instruments).

By use of a syringe pump (model sp100i2; World Precision Instruments), the arterial segments were submitted to a series of five inflation-deflation loops from 0 to 200 mmHg at a rate of 0.5 mmHg/s, serving to eliminate viscoelasticity and produce stable diameter-pressure-force loops. The inflating part of the last loop was used for the subsequent data analysis.

2.5. Material characterization

As is common in most material characterization studies (reference is made to the standard cardiovascular biomechanics textbook by Humphrey [23]), the carotid artery was considered to be a nonlinear, residually stressed, pseudoelastic, homogeneous, and orthotropic cylinder subjected to large isovolumetric deformations by lumen pressure and axial force. In the general 3D setting, circumferential and axial extension ratios $λ_{θ}$ and $λ_{z}$ were calculated relative to the zero-stress state as: $\begin{matrix} (1) & λ_{θ} = \frac{s}{S}, λ_{z} = \frac{l}{L}, \end{matrix}$ in terms of perimeter S in the zero-stress state (see Section 2.4.1. Geometrical measurements) and perimeter s in the no-load or loaded states; and in terms of ex situ and in situ lengths L and l (see Section 2.2. Euthanasia and tissue collection). Circumferential residual Green strains $E_{θ}$ were calculated at the inner and outer wall boundaries: $\begin{matrix} (2) & E_{θ} = \frac{1}{2} (λ_{θ}^{2} - 1) . \end{matrix}$

Inner radius $r_{i}$ of the carotid artery was obtained from the measured values of outer radius $r_{o}$ (see Section 2.4.2. Inflation/extension testing), and the segment’s in situ axial extension ratio $λ_{z}$ and cross-sectional area $A_{0}$ at the zero-stress state (see Section 2.4.1. Geometrical measurements), on assuming that tissue deformed isovolumetrically: $\begin{matrix} (3) & r_{i} = \sqrt{r_{o}^{2} - \frac{A_{0}}{π λ_{z}}} . \end{matrix}$ This assumption has been used in all previous studies examining the effects of estrogen on arterial mechanical properties via pressure-radius tests [9,10,12]. It was not verified for the present purposes, but the high degree of correspondence of the carotid cross-sectional area between the no-load and zero-stress states provided justification at small stresses (data not shown).

Compliance C and distensibility D to be used as structural measures were calculated from lumen area $A = π r_{i}^{2}$ and pressure P data: $\begin{matrix} (4) & C = \frac{\partial A}{\partial P}, D = \frac{1}{A} \frac{\partial A}{\partial P} . \end{matrix}$ Mean circumferential and axial Cauchy stresses $σ_{θ}$ and $σ_{z}$ were calculated as: $\begin{matrix} (5) & σ_{θ} = \frac{P r_{i}}{r_{o} - r_{i}}, σ_{z} = \frac{F + P π r_{i}^{2}}{2 π (r_{o}^{2} - r_{i}^{2})}, \end{matrix}$ where F denoted the axial force. The mean circumferential extension ratio was calculated via Eq. (1) in terms of the mid-wall vessel perimeter at the reference (zero-stress) and deformed states, given by $S = (S_{e} + S_{i}) / 2$ and $s = (s_{e} + s_{i}) / 2$ .

Qualitative description of the stress-strain curves, together with distensibility and/or compliance indices have been used by most previous investigators to assess the effects of OVX on the in vitro mechanical properties of blood vessels [8–10,13]. These indices were also used herein (viz. Eq. (4)) for comparison purposes but they are limited in that they do not permit rigorous characterization of the multiaxial material properties, unlike current microstructure-motivated models that allow proper characterization with respect to tissue microstructure. As such, we used on our inflation/extension data the neo-Hookean and four-fiber family model, proposed by Baek et al. [24]. Note the far-reaching applicability of the four-fiber family model in that it has been tested on rabbit basilar arteries [24], on mouse carotid arteries from two common models of muscular dystrophy and wild-type control [25], and for biomechanical phenotyping of carotid arteries from various genetic, pharmacological, and surgical mouse models [26]. It has also been used by our group for material characterization of various soft tissue types, e.g. porcine and rabbit carotid artery [27], porcine jugular vein [28], rabbit esophageal [29], rat large intestinal [30], and rabbit ureteral tissue [31]: $\begin{matrix} (6) & W = c (I_{1} - 3) + \sum_{m = a, d, d^{'}, c} \frac{k_{1}^{m}}{4 k_{2}^{m}} {exp [k_{2}^{m} {({(λ^{m})}^{2} - 1)}^{2}] - 1} . \end{matrix}$ Parameters c and $k_{1}^{m}$ in Eq. (6) are with stress units, $k_{2}^{m}$ are unitless, and $I_{1} = tr C$ is the first invariant of the right Cauchy–Green strain tensor C. $λ^{m} = \sqrt{n^{m} \cdot C n^{m}}$ is the extension ratio of the mth-family, whose unit vector $n^{m}$ subtends angle $a^{m}$ with the axial direction in the zero-stress state. The diagonal fibers are symmetrical, $a^{d} = - a^{d^{'}} = a$ , with equal parameters, $k_{1}^{d} = k_{1}^{d^{'}}$ , $k_{2}^{d} = k_{2}^{d^{'}}$ .

Material parameters were calculated by fitting Eq. (6) to the pressure-force-radius testing data excluding those associated with negative forces, using the Nelder–Mead algorithm in MicroCal Origin (v8.5; OriginLab Corp., Northampton, MA, USA), as in our prior studies [27–31]. The parameters were restrained by inequalities of thermodynamic origin, ensuring that the material model was convex and the parameter values physically realistic. These inequalities prescribed a zero lower limit for all parameters and an upper limit of $π / 2$ for angle a, due to the symmetry of diagonal-fiber families. The objective function minimized the normalized squared differences between the inflation/extension data and the model predictions. To ensure global extrema, optimization was repeated for various initial parameter values, covering the full operation range, i.e. ${c, k_{1}^{m}} \in [0.1, 10^{5}] Pa$ and $k_{2}^{m} \in [10^{- 3}, 10^{3}]$ , with the near-zero lower limit indicating that the contribution of a particular parameter is negligible, while $a \in [20 °, 70 °]$ to constrain the diagonal-fiber families from attaining a purely circumferential or axial orientation. The goodness of fit was estimated by the root-mean-square error ε and the determination coefficient $R^{2}$ .

2.6. Statistical methods

Results are presented as $mean \pm standard error of the mean (SEM)$ . Differences among the control, OVX, $OVX + SID$ , and $OVX + MALT$ rats were detected with one-way analysis of variance using SPSS v20.0 (SPSS Inc, Chicago, IL, USA). These were regarded as significant at the $p < 0.05$ level, in which case post-hoc multiple comparisons were performed with the Tukey test.

One specimen from the control (sample #2), OVX (sample #6), and $OVX + MALT$ (sample #6) groups was excluded from the analysis of Section 3.3. and the material characterization results of Section 3.4, even though they were submitted to inflation/extension testing, based on their force-pressure response that appeared as obvious outliers in the raw data; particularly, these specimens appeared to have been tested at above the in vivo axial extension ratio. Note, however, that they were not eliminated from the remaining analyses to which they were unrelated (Sections 3.1–3.2), and as the average results obtained and the differences between groups remained invariant by retaining them.

3. Results

3.1. Animals’ body and uterine weights, and hemodynamic measurements

There were no adverse effects by the administration of SID and/or MALT, as witnessed by monitoring body weight, fluid, and food consumption, and no pathological findings/malignancies were observed at autopsy. Water and extract consumption did not vary among groups. Body weight was similar in the four groups at baseline, but the rats of OVX groups weighed much more compared to control at euthanasia, without differences among the OVX, $OVX + SID$ , and $OVX + MALT$ groups (Table 1). The uteri of the OVX group were macroscopically atrophic and their mean weight was significantly less compared to control, but did not differ from the $OVX + SID$ and $OVX + MALT$ groups (Table 1). Pulse pressure was significantly greater in OVX compared to control rats, unlike the non-significant differences in heart rate and in systolic, diastolic, and mean blood pressures (Table 1).

Table 1
Body weight, uterine weight, and hemodynamic recordings and their statistical comparison in control, OVX, $OVX + SID$ , and $OVX + MALT$ rats

Parameter Control ( $n = 8$ ) OVX ( $n = 8$ ) $OVX + SID$ ( $n = 8$ ) $OVX + MALT$ ( $n = 8$ ) Control vs. OVX Control vs. $OVX + SID$ Control vs. $OVX + MALT$ OVX vs. $OVX + SID$ OVX vs. $OVX + MALT$ $OVX + SID$ vs. $OVX + MALT$

Body weight at baseline (g) 292.5 ± 9.2 304.4 ± 9.8 298.1 ± 8.3 301.4 ± 4.8 $NS$ $NS$ $NS$ $NS$ $NS$ $NS$

Body weight at euthanasia (g) 309.4 ± 8.1 343.8 ± 10.2 340.6 ± 5.4 342.9 ± 7.5 $p = 0.02$ $p = 0.04$ $p = 0.04$ $NS$ $NS$ $NS$

Uterine weight at euthanasia (g) 0.629 ± 0.067 0.157 ± 0.015 0.131 ± 0.005 0.161 ± 0.034 $p < 0.001$ $p < 0.001$ $p < 0.001$ $NS$ $NS$ $NS$

Systolic pressure (mmHg) 121.2 ± 1.6 124.4 ± 1.6 – – $NS$ – – – – –

Mean pressure (mmHg) 110.5 ± 1.7 109.8 ± 2.1 – – $NS$ – – – – –

Diastolic pressure (mmHg) 105.7 ± 1.8 103.0 ± 2.5 – – $NS$ – – – – –

Pulse pressure (mmHg) 15.5 ± 1.2 21.4 ± 1.8 – – $p = 0.007$ – – – – –

Heart Rate (beats/min) 332.2 ± 3.2 338.8 ± 2.9 – – $NS$ – – – – –

Parameter	Control ( $n = 8$ )	OVX ( $n = 8$ )	$OVX + SID$ ( $n = 8$ )	$OVX + MALT$ ( $n = 8$ )	Control vs. OVX	Control vs. $OVX + SID$	Control vs. $OVX + MALT$	OVX vs. $OVX + SID$	OVX vs. $OVX + MALT$	$OVX + SID$ vs. $OVX + MALT$
Body weight at baseline (g)	292.5 ± 9.2	304.4 ± 9.8	298.1 ± 8.3	301.4 ± 4.8	$NS$	$NS$	$NS$	$NS$	$NS$	$NS$
Body weight at euthanasia (g)	309.4 ± 8.1	343.8 ± 10.2	340.6 ± 5.4	342.9 ± 7.5	$p = 0.02$	$p = 0.04$	$p = 0.04$	$NS$	$NS$	$NS$
Uterine weight at euthanasia (g)	0.629 ± 0.067	0.157 ± 0.015	0.131 ± 0.005	0.161 ± 0.034	$p < 0.001$	$p < 0.001$	$p < 0.001$	$NS$	$NS$	$NS$
Systolic pressure (mmHg)	121.2 ± 1.6	124.4 ± 1.6	–	–	$NS$	–	–	–	–	–
Mean pressure (mmHg)	110.5 ± 1.7	109.8 ± 2.1	–	–	$NS$	–	–	–	–	–
Diastolic pressure (mmHg)	105.7 ± 1.8	103.0 ± 2.5	–	–	$NS$	–	–	–	–	–
Pulse pressure (mmHg)	15.5 ± 1.2	21.4 ± 1.8	–	–	$p = 0.007$	–	–	–	–	–
Heart Rate (beats/min)	332.2 ± 3.2	338.8 ± 2.9	–	–	$NS$	–	–	–	–	–

Values are $mean \pm SEM$ . $NS = non-significant$ . $Blank entry = non-available$ .

3.2. Carotid artery histology

Figures 2 and 3 show that the overall carotid artery wall structure was preserved six months after OVX. In Fig. 2 the morphometric results obtained from computerized histology are also shown. A consistent finding was the increase in all OVX groups of the cross-sectional area occupied by the media, but not of that occupied by the adventitia, leading to an increase in total wall cross-sectional area. No differences were detected in the thickness and area of the various layers and the entire wall among the OVX groups, as well as in the inner and outer diameters of the carotid artery between all OVX groups and control.

As remarked in Fig. 3 summarizing the composition parameters measured, the relative area density of collagen in tunica media and adventitia, quantified in Sirius-red sections, was greater in all OVX groups than control. So was elastin area density in the media, quantified in orcein sections, while cell area density in the media and adventitia was decreased. As a result, elastin and collagen area densities in the entire wall were greater in OVX arteries than control, whereas cell area density was decreased. Unchanged was the number of lamellar units in the media and the number of nuclear profiles per unit surface area in all layers and the entire wall (data not shown). These data imply that arterial growth in response to OVX was characterized by collagen and elastin buildup.

3.3. Carotid artery mechanical properties

3.3.1. Geometrical parameters

Table 2 lists the morphological parameters measured from the digitized photographs of the carotid artery segments in the no-load and zero-stress state. In the no-load state, the inner diameter did not vary between control and OVX, $OVX + SID$ , or $OVX + MALT$ rats, whereas the outer diameter was significantly larger in the latter groups, by virtue of the greater wall thickness; consequently, the ratio of wall thickness to inner diameter was significantly higher. Wall thickness did not differ significantly between the no-load and zero-stress states (data not shown).

Fig. 2.

Characteristic low-magnification histological sections of the common carotid artery stained with hematoxylin-eosin from (A) a control, (B) an OVX, (C) an $OVX + SID$ , and (D) an $OVX + MALT$ rat. The scale bar refers to all the histological sections. Histomorphometric parameters ((E) inner and outer diameter; (F) media, adventitia, and entire wall thickness; (G) media, adventitia, and entire wall area) of the common carotid artery and their statistical comparison in control, OVX, $OVX + SID$ , and $OVX + MALT$ rats. Values are $mean \pm SEM$ . Significant differences are only shown.

Fig. 3.

High-magnification histological sections of the common carotid artery, adjacent to those of Fig. 1, stained with orcein for elastin ((A)–(D)) and sirius-red for collagen ((E)–(H)) differentiation from ((A), (E)) a control, an OVX ((B), (F)), an $OVX + SID$ ((C), (G)), and an $OVX + MALT$ rat ((D), (H)). The scale bar refers to all the histological sections. Compositional parameters ((I) elastin, (J) collagen, and (K) cells area densities) of media, adventitia, and the entire wall of the carotid artery and their statistical comparison in control, OVX, $OVX + SID$ , and $OVX + MALT$ rats. Values are $mean \pm SEM$ . Significant differences are only shown.

The opening angle and residual extension ratio results at the intimal and adventitial surfaces of the carotid artery are depicted in Table 2. Opening angle was significantly higher in $OVX + SID$ and $OVX + MALT$ than control rats, unlike the compressive intimal residual extension ratio and the tensile adventitial residual extension ratio that did not vary. The axial extension ratio of the carotid artery at the in situ state was significantly lower in OVX, $OVX + SID$ , and $OVX + MALT$ than control rats. There were no differences in any of the abovementioned parameters among the OVX groups.

3.3.2. Inflation/extension data

Figures 4 and 5 illustrate the findings from inflation/extension testing on the common carotid arteries at the in situ extension ratio and under passive smooth muscle that permit direct comparison among groups at their individual in situ condition. All arteries exhibited a bi-phasic deformational response (Figs 4(A) and (C)), with a low-pressure phase (0–80 mmHg) referring to a great variation of outer radius and wall thickness, and a physiologic and supra-physiologic phase referring to radius and thickness locking. Note the quite similar curve shape of OVX and control arteries. Pressure inflation in the former led to a somewhat greater radius enlargement at all levels of pressure that was associated with an equally compliant and distensible wall than control, however; Figs 5(A) and (B). Quite similar were also the cumulative circumferential and axial stress-circumferential extension ratio data between the OVX and control carotids shown in Figs 5(C) and (D), noting that the data referred to a significantly smaller axial in situ extension ratio in OVX compared to control rats (Table 1). Note also the striking similarity in the curves of OVX, $OVX + SID$ , and $OVX + MALT$ rats, suggesting that SID and MALT administration did not influence the mechanical response of OVX arteries.

3.4. Carotid artery material characterization

Pressure-radius-force data and curve-fits are shown in Fig. 6 for all the carotid arteries tested. The neo-Hookean and four-fiber family model provided a realistic characterization of the inflation-extension response ( $ε \sim 0.2$ ), by properly capturing the entire dataset. Table 3 summarizes the average model parameters optimized, along with statistical comparisons among the rat groups. Compared to control, parameters $k_{1, 2}^{a}$ and $k_{1}^{d}$ , specifying the physiologic and high-pressure wall stiffness, i.e. the collagen-fiber contribution, were greater in the OVX, $OVX + SID$ , and $OVX + MALT$ groups unlike the orientation angle a of diagonal fibers and parameter c, specifying the low-pressure wall stiffness, i.e. the contribution of elastin and passive muscle, that were smaller; quite similar among groups were the values of parameter $k_{1, 2}^{c}$ and $k_{2}^{d}$ . Despite the non-significant differences of parameters likely due to their great scatter within each group, the Cauchy stress predictions by the model at equibiaxial stretch states were significantly higher in the OVX, $OVX + SID$ , and $OVX + MALT$ groups than control especially in the axial direction (Fig. 7), reflecting that the carotid artery in the OVX groups was stiffer than control at physiologic to higher pressures. Distinct from control, stress was predicted in all OVX groups as higher axially than circumferentially, indicative of certain anisotropy.

Table 2
Geometrical parameters in the no-load and zero-stress state of the common carotid artery and their statistical comparison in control, OVX, $OVX + SID$ , and $OVX + MALT$ rats

Parameter Control ( $n = 7$ ) OVX ( $n = 7$ ) $OVX + SID$ ( $n = 8$ ) $OVX + MALT$ ( $n = 7$ ) Control vs. OVX Control vs. $OVX + SID$ Control vs. $OVX + MALT$ OVX vs. $OVX + SID$ OVX vs. $OVX + MALT$ $OVX + SID$ vs. $OVX + MALT$

NLT (mm) 0.104 ± 0.004 0.120 ± 0.003 0.119 ± 0.003 0.123 ± 0.004 $p = 0.04$ $p = 0.03$ $p = 0.003$ $NS$ $NS$ $NS$

NLID (mm) 0.838 ± 0.022 0.802 ± 0.016 0.808 ± 0.011 0.799 ± 0.018 $NS$ $NS$ $NS$ $NS$ $NS$ $NS$

$NLT / NLID$ (–) 0.126 ± 0.006 0.150 ± 0.004 0.147 ± 0.003 0.157 ± 0.008 $p = 0.02$ $p = 0.03$ $p = 0.001$ $NS$ $NS$ $NS$

$OA$ (deg) 118.2 ± 5.2 136.2 ± 5.4 135.8 ± 4.4 140.5 ± 5.4 $NS$ $p = 0.05$ $p = 0.03$ $NS$ $NS$ $NS$

IRER (–) 0.972 ± 0.013 0.934 ± 0.010 0.950 ± 0.010 0.937 ± 0.015 $NS$ $NS$ $NS$ $NS$ $NS$ $NS$

ORER (–) 1.113 ± 0.028 1.170 ± 0.021 1.151 ± 0.017 1.172 ± 0.031 $NS$ $NS$ $NS$ $NS$ $NS$ $NS$

AER (–) 1.470 ± 0.046 1.256 ± 0.060 1.259 ± 0.024 1.241 ± 0.039 $p = 0.009$ $p = 0.008$ $p = 0.005$ $NS$ $NS$ $NS$

Parameter	Control ( $n = 7$ )	OVX ( $n = 7$ )	$OVX + SID$ ( $n = 8$ )	$OVX + MALT$ ( $n = 7$ )	Control vs. OVX	Control vs. $OVX + SID$	Control vs. $OVX + MALT$	OVX vs. $OVX + SID$	OVX vs. $OVX + MALT$	$OVX + SID$ vs. $OVX + MALT$
NLT (mm)	0.104 ± 0.004	0.120 ± 0.003	0.119 ± 0.003	0.123 ± 0.004	$p = 0.04$	$p = 0.03$	$p = 0.003$	$NS$	$NS$	$NS$
NLID (mm)	0.838 ± 0.022	0.802 ± 0.016	0.808 ± 0.011	0.799 ± 0.018	$NS$	$NS$	$NS$	$NS$	$NS$	$NS$
$NLT / NLID$ (–)	0.126 ± 0.006	0.150 ± 0.004	0.147 ± 0.003	0.157 ± 0.008	$p = 0.02$	$p = 0.03$	$p = 0.001$	$NS$	$NS$	$NS$
$OA$ (deg)	118.2 ± 5.2	136.2 ± 5.4	135.8 ± 4.4	140.5 ± 5.4	$NS$	$p = 0.05$	$p = 0.03$	$NS$	$NS$	$NS$
IRER (–)	0.972 ± 0.013	0.934 ± 0.010	0.950 ± 0.010	0.937 ± 0.015	$NS$	$NS$	$NS$	$NS$	$NS$	$NS$
ORER (–)	1.113 ± 0.028	1.170 ± 0.021	1.151 ± 0.017	1.172 ± 0.031	$NS$	$NS$	$NS$	$NS$	$NS$	$NS$
AER (–)	1.470 ± 0.046	1.256 ± 0.060	1.259 ± 0.024	1.241 ± 0.039	$p = 0.009$	$p = 0.008$	$p = 0.005$	$NS$	$NS$	$NS$

Values are $mean \pm SEM$ . $NLT = no-load thickness$ ; $NLID = no-load inner diameter$ ; $OA = opening angle$ ; $IRER = inner (intimal) residual extension ratio$ ; $ORER = outer (adventitial) residual extension ratio$ ; $AER = axial extension ratio in situ$ ; $NS = non-significant$ .

Fig. 4.

(A) Outer radius, (B) axial force, and (C) wall thickness vs. lumen pressure data of control, OVX, $OVX + SID$ , and $OVX + MALT$ carotid arteries at their in situ axial extension ratio and under passive smooth muscle. Data points were averaged at specific values of lumen pressure from 0 to 200 mmHg in steps of 10 mmHg.

Fig. 5.

(A) Compliance and (B) distensibility vs. lumen pressure data, and (C) circumferential and (D) axial stress vs. circumferential extension ratio data of control, OVX, $OVX + SID$ , and $OVX + MALT$ carotid arteries at their in situ axial extension ratio and under passive smooth muscle. Data points were averaged at specific values of lumen pressure from 0 to 200 mmHg in steps of 10 mmHg.

Fig. 6.

Experimental lumen pressure (upper panel) and axial force (lower panel) data as a function of outer radius for all carotid arteries tested and fits by the neo-Hookean and four-fiber family model for (A) the control, (B) OVX, (C) $OVX + SID$ , and (D) $OVX + MALT$ groups. Data are shown in steps of 10 mmHg for clarity. The numbers at the end of each pressure-radius and force-radius curve are shown to associate curves with specimens from each group. Note that specimen #2 from the control group, specimen #6 from the OVX group, and specimen #6 from the $OVX + MALT$ group shown above were excluded from the material characterization results of Table 3, since their force-pressure response appeared as outliers in the raw data. Note too that for samples #3–5 from the $OVX + MALT$ group, the model was fitted to the pressure-force-radius dataset referring to strictly positive forces, i.e. the fitted line did not cover the high-pressure range associated with negative forces and buckling.

4. Discussion

4.1. Animal models and general considerations

A number of studies [8–14] have compared the mechanical behavior of arteries from sham-operated and OVX animals before and after hormone replacement therapy, but relatively young immature animals were only examined, even though the mechanical properties of arteries from common laboratory animals are modified with aging, as do the hemodynamic conditions, and most importantly the capacity of blood vessels to remodel may be impaired by age [32–34]. To overcome this shortcoming, OVX was performed in the present study to middle-aged rats, thus ascertaining whether the arterial remodeling is age-related. Particularly, we selected mature rats at the age of 10 months, since this is the age that they achieve peak bone mass and is considered most appropriate for modeling postmenopausal osteoporosis in women [35].

Inherent differences among the various studies may be responsible for the inconsistencies in the reported remodeling for young animals, such as lumen narrowing in response to depleted estrogen levels in some cases [9,10] and increased wall thickness in other cases [12], and stiffening [12,14] unlike the invariant [8–10,13] and decreased stiffness [11] disclosed by other authors. The reasons behind these discrepancies are uncertain but may be ascribed to differences in the duration of OVX (few weeks [8,10–12] vs. several months [9,13]) or in the blood vessels tested (large elastic [11–13] vs. small resistance arteries [8–10]). Other likely explanations relate to the methods employed to determine the mechanical properties (in vitro [8–10,13] vs. in vivo [11,12,14]) and dimensions (no-load [8] vs. loaded state [9–12]). Despite the rather incomplete data, Section 4.2–4.4 lay emphasis on interesting analogies between the present and past studies, while providing a comprehensive assessment of the mechanical, morphological, and structural aspects of remodeling.

Table 3
Parameters and fit quality of the four-fiber family model for the common carotid artery and their statistical comparison in control, OVX, $OVX + SID$ , and $OVX + MALT$ rats

Parameter Control ( $n = 7$ ) OVX ( $n = 7$ ) $OVX + SID$ ( $n = 8$ ) $OVX + MALT$ ( $n = 7$ ) Control vs. OVX Control vs. $OVX + SID$ Control vs. $OVX + MALT$ OVX vs. $OVX + SID$ OVX vs. $OVX + MALT$ $OVX + SID$ vs. $OVX + MALT$

$c (kPa)$ 30.318 ± 10.701 4.905 ± 3.134 14.743 ± 5.972 1.295 ± 1.295 $p = 0.04$ $NS$ $p = 0.02$ $NS$ $NS$ $NS$

$k_{1}^{a} (kPa)$ 9.772 ± 4.757 51.602 ± 19.652 32.421 ± 10.336 38.882 ± 23.099 $NS$ $NS$ $NS$ $NS$ $NS$ $NS$

$k_{2}^{a} (–)$ 0.443 ± 0.186 0.852 ± 0.445 0.992 ± 0.643 2.010 ± 0.670 $NS$ $NS$ $NS$ $NS$ $NS$ $NS$

$k_{1}^{d} (kPa)$ 27.488 ± 12.677 55.178 ± 13.757 56.116 ± 20.529 69.353 ± 20.002 $NS$ $NS$ $NS$ $NS$ $NS$ $NS$

$k_{2}^{d} (–)$ 1.608 ± 0.435 1.850 ± 1.395 2.138 ± 1.037 0.938 ± 0.400 $NS$ $NS$ $NS$ $NS$ $NS$ $NS$

$k_{1}^{c} (kPa)$ 4.799 ± 4.404 3.486 ± 2.245 11.952 ± 5.285 7.018 ± 2.906 $NS$ $NS$ $NS$ $NS$ $NS$ $NS$

$k_{2}^{c} (–)$ 1.467 ± 0.464 2.055 ± 0.776 2.392 ± 1.162 2.482 ± 0.656 $NS$ $NS$ $NS$ $NS$ $NS$ $NS$

$a (rad)$ 0.879 ± 0.038 0.840 ± 0.039 0.797 ± 0.057 0.766 ± 0.071 $NS$ $NS$ $NS$ $NS$ $NS$ $NS$

$ε (–)$ 0.165 ± 0.029 0.222 ± 0.027 0.180 ± 0.025 0.247 ± 0.016 – – – – – –

$R^{2} (–)$ 0.947 ± 0.015 0.922 ± 0.015 0.949 ± 0.012 0.897 ± 0.013 – – – – – –

Parameter	Control ( $n = 7$ )	OVX ( $n = 7$ )	$OVX + SID$ ( $n = 8$ )	$OVX + MALT$ ( $n = 7$ )	Control vs. OVX	Control vs. $OVX + SID$	Control vs. $OVX + MALT$	OVX vs. $OVX + SID$	OVX vs. $OVX + MALT$	$OVX + SID$ vs. $OVX + MALT$
$c (kPa)$	30.318 ± 10.701	4.905 ± 3.134	14.743 ± 5.972	1.295 ± 1.295	$p = 0.04$	$NS$	$p = 0.02$	$NS$	$NS$	$NS$
$k_{1}^{a} (kPa)$	9.772 ± 4.757	51.602 ± 19.652	32.421 ± 10.336	38.882 ± 23.099	$NS$	$NS$	$NS$	$NS$	$NS$	$NS$
$k_{2}^{a} (–)$	0.443 ± 0.186	0.852 ± 0.445	0.992 ± 0.643	2.010 ± 0.670	$NS$	$NS$	$NS$	$NS$	$NS$	$NS$
$k_{1}^{d} (kPa)$	27.488 ± 12.677	55.178 ± 13.757	56.116 ± 20.529	69.353 ± 20.002	$NS$	$NS$	$NS$	$NS$	$NS$	$NS$
$k_{2}^{d} (–)$	1.608 ± 0.435	1.850 ± 1.395	2.138 ± 1.037	0.938 ± 0.400	$NS$	$NS$	$NS$	$NS$	$NS$	$NS$
$k_{1}^{c} (kPa)$	4.799 ± 4.404	3.486 ± 2.245	11.952 ± 5.285	7.018 ± 2.906	$NS$	$NS$	$NS$	$NS$	$NS$	$NS$
$k_{2}^{c} (–)$	1.467 ± 0.464	2.055 ± 0.776	2.392 ± 1.162	2.482 ± 0.656	$NS$	$NS$	$NS$	$NS$	$NS$	$NS$
$a (rad)$	0.879 ± 0.038	0.840 ± 0.039	0.797 ± 0.057	0.766 ± 0.071	$NS$	$NS$	$NS$	$NS$	$NS$	$NS$
$ε (–)$	0.165 ± 0.029	0.222 ± 0.027	0.180 ± 0.025	0.247 ± 0.016	–	–	–	–	–	–
$R^{2} (–)$	0.947 ± 0.015	0.922 ± 0.015	0.949 ± 0.012	0.897 ± 0.013	–	–	–	–	–	–

${c, k_{1}^{a}, k_{2}^{a}, k_{1}^{d}, k_{2}^{d}, k_{1}^{c}, k_{2}^{c}, a} = model parameters$ ; $ε = root-mean-square error of curve-fitting$ ; $R^{2} = determination coefficient$ .

Values are $mean \pm SEM$ . $NS = non-significant$ . $Blank entry = no statistical analysis performed$ .

4.2. Mechanical/morphological changes of arterial response to OVX: Comparison to previous studies

The outer radius-lumen pressure curves of the carotid arteries from all three OVX groups were displaced to greater radii (Fig. 4(A)), caused by their greater wall thickness (Fig. 4(C)) and inner radius. This finding may be expected based on the earliest mechanics-based study by Cox and Fischer [36], who conducted pressure-radius measurements under inflation at in vivo length in the presence of smooth muscle inhibitors. The curves for carotid arteries treated with estradiol were shifted to smaller radii, consistent with the effect of estrogen in decelerating the weight gain of rats (Table 1), and those for testosterone-treated arteries to greater radii, with qualitatively similar results in castrated and non-castrated male rats. Three-week administration of estradiol decreased the incremental elastic modulus, a measure of material stiffness, whereas administration of testosterone increased it, albeit to a lesser extent, through the opposite effects of these steroids on the elastin/collagen accumulation in the arterial wall that was demonstrated in their companion study [37].

Fig. 7.

Carotid artery stiffness comparisons and anisotropy assessment via the prediction of (A) circumferential and (B) axial stresses at equibiaxial extension ratio states by the neo-Hookean and four-fiber family model for the control, OVX, $OVX + SID$ , and $OVX + MALT$ groups.

Acs et al. [9 ,10] conducted in vitro tests of saphenous arteries from OVX and control rats at in vivo length and potential modification in vasoactivity was appropriately considered. Contrary to our findings, 12-week pharmacological OVX shifted the passive pressure-radius curves to smaller inner radii that were restored by hormone replacement therapy, with unchanged vessel wall thickness. No significant changes in distensibility (structural) and incremental elastic modulus (material index) were found, in certain analogy to the invariant distensibility as well as circumferential and axial stress vs. circumferential extension ratio curves for the OVX and control rats reported herein (Fig. 5). More recently, Guo et al. [12] reported data from in vivo pressure-radius experiments on mouse aorta, and carotid and femoral arteries. Much like the present data, they disclosed that surgical OVX for 4 weeks increased wall thickness but not lumen radius and, although it did not affect the axial stiffness as we found, it substantially increased the circumferential stiffness. Unfortunately, axial force was not measured in any of the abovementioned studies, so that the pressure-radius data alone constitute in essence uniaxial data that do not suffice for multiaxial material characterization. Other important information that is missing regards the in vivo axial prestretches of vessels, and their zero-stress state was not examined except in [12] but no opening angle and/or residual strain data were presented.

Park et al. [13] reported that estrogen deficiency did not have an effect on aortic stiffness of rats at 12 weeks after surgical OVX, basing their conclusions on ring tests. Their finding of unchanged elastic modulus, and ultimate tensile stress and stretch were corroborated by their biochemical data, that is the absence of calcium deposition and elastin content change; the latter finding being at variance with our light densitometry data (Fig. 3). Note, however, that ring tests are less preferred for both experimental and theoretical reasons, as they provide uniaxial data with decoupled circumferential and axial loads as experienced physiologically, the unloaded state of tissue is considered as its zero-stress state, and bending during the initial testing phases is ignored. Similar reservations, therefore, hold for the ring studies by Hansen et al. [8] on rabbit cerebral and coronary arteries that disclosed unchanged elastic modulus at 6 weeks after surgical OVX, as well as unchanged collagen content and morphologic features, i.e. lumen diameter, wall thickness, and layer thicknesses.

Other parameters being alike in two available pulse wave velocity studies, i.e. unchanged wall thickness and lumen diameter (unlike our results in Table 2), the significant aortic pressure rise disclosed in [14] after 2 months may be responsible for the increased pulse wave velocity in OVX rats. Even adjusting for pressure, this finding contrasts with [11] showing a reduced pulse wave velocity and elastic modulus after a 3-week OVX period, despite the unchanged pressure. However, inferences of aortic stiffness from pulse wave velocity measurements are not straightforward given the complexity and number of their determinants, including wall dimensions, passive wall elements, smooth muscle tone, and the level of distending pressure.

4.3. Material characterization

The mechanical remodeling of the carotid artery in OVX rats was mathematically assessed by characterizing our inflation-extension data using a microstructure-motivated material model (Eq. (6)). As such, we employed the neo-Hookean and four-fiber family model that is extensively used in the literature and permitted realistic data characterization for both OVX and control arteries (Fig. 6). Although non-significant due to the wide scatter, the parameter values optimized exhibited marked differences between OVX and control arteries, i.e. a decrease in the neo-Hookean function parameter (associated with elastin and other non-collagenous elements) and an increase in several four-fiber family parameters (associated with collagen fibers; Table 3). At equibiaxial stretch states, the model predicted a significantly stiffer carotid artery wall in all OVX groups compared to control (Fig. 7), justified by the increased medial and adventitial collagen contents found that are circumferentially- and axially-aligned, respectively [27]. Note, however, that the material characterization results would have been more robust had the inflation testing been performed at several and not only at the in situ axial extension ratio that was unfortunately performed in the current experiments. Comprehensive testing protocols, including a large variety of loading scenarios, are necessary for accurate parameter identification of nonlinear material models. Consideration of pressure-radius-force data at additional axial extensions and of axial force-length test data at different pressures, as done in the early study for the rat carotid artery by Weizsäcker et al. [38], and more recently for the mouse carotid artery by Gleason et al. [25] and Bersi et al. [26], would have considerably strengthened the descriptive and predictive capabilities of the chosen material model.

Humphrey et al. [39] (see also [26]) suggested the determination of the in situ axial extension ratio from the force-length relationship, where various isobaric axial force-length curves intersect, or alternatively from the force-pressure relationship, where force remains nearly constant; especially when direct measurements are difficult, for instance with small animals since measurements may be affected by changes in body position. Our average and individual data for the control group in Figs 4(B) and 6 are consistent with such a definition, as we found that the axial force was nearly unchanged with pressure inflation at the directly-measured in situ value of axial extension ratio, while the somewhat negative slopes of the average force-pressure data for the OVX groups, and especially of individual data, indicated a testing condition below the in situ axial extension ratio as defined by a constant force-pressure relationship. Given also the difficulties in interpreting our data, due to the absence of testing data at multiple axial extensions, the presented comparisons among groups should be seen with caution.

Notwithstanding this limitation, the observation of reduced in situ axial extension ratio in OVX rats (Table 2) is valuable and suggests that the physiologic environment of carotid arteries changed in response to OVX, so that the circumferential and axial stress-extension ratio responses of the control and all OVX groups be quite similar when examined at their individual in situ axial extension ratios (Fig. 5). Humphrey et al. [39] examined the biomechanical adaptation in diverse conditions, i.e. modified cell-matrix interactions in muscular dystrophy, modified matrix composition in fibrillin-1 deficiency, modified pulse pressure in aortic banding, and modified axial extension in organ culture, and found that mouse common carotid arteries modified drastically their axial extension ratio to normalize the overall stress state toward the homeostatic condition. Our results corroborate these important observations and add confirmation to the statement made by those authors that “Despite these different types of perturbation from normal, the arterial wall compensates in each case via a reduction in axial stretch, that is, an increase in unloaded length.” [39].

4.4. Structural and zero-stress state changes

The carotid artery suffered thickening, without change in lumen diameter, leading to an increase in their ratio and in wall cross-sectional area (Table 2 and Fig. 2). These findings are in line with the observations in [37] that carotid arteries from estradiol-treated rats had a significantly smaller wall area; these authors also reported that arteries from castrated rats had a significantly greater ratio of wall thickness to lumen radius compared to control rats. We additionally observed that the different wall layers grew at dissimilar rates, with the carotid arteries of OVX rats exhibiting a thicker media but not adventitia relative to control rats (Fig. 2).

The geometrical remodeling of tunica media in OVX rats was accompanied by an increase in collagen and elastin; similar was the trend for collagen in the adventitia. This similarity together with the growth of media led to increased collagen and elastin total wall percentage contents in OVX rats (Fig. 3). In most respects, these findings are not particularly unexpected; they seem consistent with early studies showing that exogenous estradiol treatment reversed the aortic growth in OVX female rodents or other animal models caused by the accumulation of elastin, collagen [40–42], and glycosaminoglycans [43], irrespective of body weight changes. Analogous results were more recently found in a human cell culture study [44] examining the direct effects of sex steroids devoid of their hemodynamic effects that are present in animal models. On the other hand, the initial age of OVX animals in the abovementioned studies was much less than that of the animals in the present report that more closely reflects the age of postmenopausal women.

Dobrin et al. [45] proposed that the axial retractive force in large arteries is primarily exerted by elastin, which is the main view held today. Elastin content was not reduced or fragmented in OVX rats to justify the significantly reduced axial extension ratio found herein (Table 2). This may rather be ascribed to arterial growth effected in both the circumferential and axial directions, as proposed by Langille et al. [46]. The in situ axial extension ratio we report about the control rats was considerably smaller than the 1.7-value reported in the literature; reference is made to the data by Weizsäcker et al. [38] on 9-month old rats and our recent study [22] on 6-month old rats. However, female rats of very large age were presently examined, i.e. 16-month-old, so that the $\sim 1.5$ -value found is consistent with the age-related decrease in axial retraction of various human arteries upon excision that has been documented by Learoyd and Taylor [47]. These authors measured the amount of vessel shortening as a percentage of the in situ length, reporting for the carotid artery a decrease from 26% to 18% between young (11–20 years) and middle-aged (36–52 years) subjects. We are not aware of previous studies examining the in situ axial extension ratio of the carotid or other central artery in aged rats to compare with the present findings, but the marked decrease may again be attributed to the age-related arterial growth, requiring a smaller axial extension ratio to fit a larger artery to a fixed length inside the body.

As a measure of the zero-stress state, we found that the opening angle varied with the morphological and structural remodeling. Specifically, the increased opening angle of the carotid artery in response to OVX associates directly with the increased ratio of wall thickness to lumen diameter (Table 2). It also corroborates Fung’s hypothesis [48], that is of the inner wall layers undergoing greater growth than the outer, as we found the media of the carotid artery from OVX rats to be significantly thicker and to occupy a greater cross-sectional area than that of control rats, unlike the adventitia that was equally thick (Fig. 2); noting that a reliable thickness/area measurement of the adventitia with its highly irregular structure around the circumference was not easy, and hence more likely to be afflicted with errors. Opening angle changes may also be attributed to changes in material properties, e.g. [49]. The increased opening angle in OVX rats may thus be associated to the stiffer carotid artery wall (Fig. 7). The consideration of age-related differences between the present and our previous control data on younger rats adds confirmation to well-known geometrical changes occurring with aging in the vasculature. Particularly, it discloses an age-related increase in opening angle (cf. $118.2 \pm 5.2 deg$ herein (Table 2) with $93.2 \pm 7.6 deg$ in Table 2 of [22]) and a minor increase in carotid dimensions (c.f. the histologically-determined inner radius and wall thickness shown in Fig. 2 herein with those in Fig. 3 of [22] and the pressure-radius relationship in Fig. 4 herein with that in Fig. 8 of [22]), as has been demonstrated by Badreck-Amoudi et al. [33] for the female rat aorta.

4.5. Consideration of the effects of SID

It has been observed that patient compliance to some of the current available treatments for osteoporosis, such as hormone replacement therapy and bisphosphonates is low, due to actual or fear of adverse effects [50,51]. Many patients therefore, in search of alternative protective/therapeutic options, resort to natural substances or ethnic functional foods [52,53]. Studies of the plant extracts that are being consumed to this end have demonstrated increased isoflavone and other compound levels, which have an affinity for estrogen receptors [54,55] and may thereby explain their bone-sparing effects during estrogen deficiency.

The extract from the Mediterranean plant Sideritis euboea that was used in the present study has been shown to enhance osteoblast proliferation in vitro [16], as well as to have a protective effect on bone mineral density and strength of OVX rats [17]. Unfortunately, this study does not report a beneficial effect as regards the structure, morphology, and mechanical properties of a large elastic artery, which is evidenced by the non-significant differences in any of these properties among the OVX, $OVX + SID$ , and $OVX + MALT$ groups.

4.6. Consideration of the hemodynamic aspects of arterial remodeling

Estrogen has a number of effects on cardiovascular function and can regulate the mechanical properties of arteries either by direct effects on the vascular cells or indirectly by systemic effects, so that it is important to identify the stimulus for the observed changes; i.e. is it a systemic action on, e.g. hemodynamics, or a direct action on smooth muscle cells? Furthermore, are the effects of OVX on arterial wall mechanical properties, morphology, and structure mediated by a physical (i.e. change in pressure/flow) or a biochemical stimulus (i.e. change in NO production) with estrogen deficiency? Systolic, diastolic, and mean blood pressures measured in the tail artery of OVX rats, which need to be seen with caution, did not differ from control whereas the increased pulse pressure may not be sufficient to explain a consistent hypertensive effect per se as it’s magnitude was not in the hypertensive range [56,57] and not likely to be great enough to induce the observed remodeling in OVX rats. Although we have no data on cardiac output, it is unlikely that it was much depressed in the absence of bradycardia that is on account of a reduced stroke volume and a previous investigation disclosed no variations in hindquarters and carotid blood flow in OVX rats [11]. We could also not study pressure variations among all groups, because the data from the $OVX + SID$ and $OVX + MALT$ groups were accidentally lost. Interestingly and despite the non-significant differences in arterial pressures measured herein and in most previous reports [9–13] except [14], the observed remodeling is very much alike with that caused by a blood pressure rising effect, as would be consistent with the unchanged lumen diameter, the increase in wall thickness and particularly of its inner layers, in opening angle, and in stiffness caused by collagen and elastin deposition, and the decrease in the in situ axial extension ratio [56,57].

4.7. Concluding remarks

Estrogen depletion in postmenopausal women can produce cardiovascular system changes that can range from hypertension, atherosclerosis, arrhythmias up to cardiac arrest or stroke [58,59]. Animal studies are needed to determine the detrimental effects of depleted estrogen levels in the cardiovascular system and as such the present communication reports morphological, histological, and mechanical findings for the common carotid artery from middle-aged control (sham-operated) and OVX rats. Additional studies are clearly needed to provide more robust inflation/extension data for material characterization and to better define the time course of arterial changes following OVX, thereby offering important insight into the effects of estrogen on vascular structure and function.

References

Barrett-Connor

. Sex differences in coronary artery disease: Why are women so superior? The 1995 ancel keys lecture. Circulation. 1997;95:252–264.

Mendelsohn

Karas

. Molecular and cellular basis of cardiovascular gender differences. Science. 2005;38:1583–1587. doi:10.1126/science.1112062.

Murphy

. Estrogen signaling and cardiovascular disease. Circ Res. 2011;109:687–696. doi:10.1161/CIRCRESAHA.110.236687.

Knowlton

Lee

. Estrogen and the cardiovascular system. Pharmacol Ther. 2012;135:54–70. doi:10.1016/j.pharmthera.2012.03.007.

Waddell

Dart

Gatzka

Cameron

Kingwell

. Women exhibit a greater age-related increase in proximal aortic stiffness than men. J Hypertens. 2001;19:2205–2212. doi:10.1097/00004872-200112000-00014.

Berry

Cameron

Dart

Dewar

Gatzka

Jennings

, et al. Large-artery stiffness contributes to the greater prevalence of systolic hypertension in elderly women. J Am Geriatr Soc. 2004;52:368–373. doi:10.1111/j.1532-5415.2004.52107.x.

Rajkumar

Kingwell

Cameron

Waddell

Mehra

Christophidis

, et al. Hormonal therapy increases arterial compliance in postmenopausal women. J Am Coll Cardiol. 1997;30:350–356. doi:10.1016/S0735-1097(97)00191-5.

Hansen

Skajaa

Aalkjaer

Oxlund

Glavind

Petersen

, et al. The effect of oophorectomy on mechanical properties of rabbit cerebral and coronary isolated small arteries. Am J Obstet Gynecol. 1996;175:1272–1280. doi:10.1016/S0002-9378(96)70040-6.

Ács

Székács

Nádasy

Várbíró

Kakucs

Monos

. The effect of ovariectomy and oestrogen replacement on small artery biomechanics in the rat. Br J Obstet Gynaecol. 1999;106:148–154. doi:10.1111/j.1471-0528.1999.tb08215.x.

10.

Ács

Székács

Nádasy

Várbíró

Miklós

Szentivanyi

Jr , et al. Effects of combined sex hormone replacement therapy on small artery biomechanics in pharmacologically ovariectomized rats. Maturitas. 2000;34:83–92. doi:10.1016/S0378-5122(99)00086-9.

11.

Tatchum-Talom

Martel

Marette

. Influence of estrogen on aortic stiffness and endothelial function in female rats. Am J Physiol Heart Circ Physiol. 2002;282:H491–H498. doi:10.1152/ajpheart.00589.2001.

12.

Guo

Ren

Levin

Kassab

. Estrogen modulates the mechanical homeostasis of mouse arterial vessels through nitric oxide. Am J Physiol Heart Circ Physiol. 2006;290:H1788–H1797. doi:10.1152/ajpheart.01070.2005.

13.

Park

Omi

Nosaka

Kitajima

Ezawa

. Estrogen deficiency and low-calcium diet increased bone loss and urinary calcium excretion but did not alter arterial stiffness in young female rats. J Bone Miner Metab. 2008;26:218–225. doi:10.1007/s00774-007-0822-4.

14.

Losurdo

Grillo

Panizon

Cappellari

Fabris

Bardelli

, et al. Supplementation of omega-3 polyunsaturated fatty acids prevents increase in arterial stiffness after experimental menopause. J Cardiovasc Pharmacol Ther. 2014;19:114–120. doi:10.1177/1074248413500716.

15.

Nichols

O’Rourke

Vlachopoulos

. McDonald’s blood flow in arteries. Theoretical, experimental and clinical principles. 6th ed. London: Hodder Arnold; 2011.

16.

Kassi

Paliogianni

Dontas

Aligiannis

Halabalaki

Papoutsi

, et al. Effects of Sideritis euboea (Lamiaceae) aqueous extract on IL-6, OPG and RANKL secretion by osteoblasts. Nat Prod Commun. 2011;6:1689–1696.

17.

Dontas

Lelovas

Kourkoulis

Aligiannis

Paliogianni

Mitakou

, et al. Protective effect of Sideritis euboea extract on bone mineral density and strength of ovariectomized rats. Menopause. 2011;18:915–922. doi:10.1097/gme.0b013e31820ce580.

18.

Borrelli

Ernst

. Alternative and complementary therapies for the menopause. Maturitas. 2010;66:333–343. doi:10.1016/j.maturitas.2010.05.010.

19.

Shen

Chyu

Wang

. Tea and bone health: Steps forward in translational nutrition. Am J Clin Nutr. 2013;98:1694S–1699S.

20.

Kritharis

Kakisis

Giagini

Manos

Stergiopulos

Tsangaris

, et al. Biomechanical, morphological and zero-stress state characterization of jugular vein remodeling in arteriovenous fistulas for hemodialysis. Biorheology. 2010;47:297–319. doi:10.3233/BIR-2011-0578.

21.

Kritharis

Giagini

Kakisis

Dimitriou

Stergiopulos

Tsangaris

, et al. Time course of flow-induced adaptation of carotid artery biomechanical properties, structure, and zero-stress state in the arteriovenous shunt. Biorheology. 2012;49:65–82. doi:10.3233/BIR-2012-0602.

22.

Lampropoulos

Sokolis

. Large artery biomechanical, geometrical, and structural remodeling elicited by long-term propranolol administration in an animal model. Biorheology. 2016;53:151–170. doi:10.3233/BIR-16090.

23.

Humphrey

. Cardiovascular solid mechanics: Cells, tissues, and organs. New York: Springer; 2002.

24.

Baek

Gleason

Rajapogal

Humphrey

. Theory of small on large: Potential utility in computations of fluid-solid interactions in arteries. Comput Methods Appl Mech Engrg. 2007;196:3070–3078. doi:10.1016/j.cma.2006.06.018.

25.

Gleason

Dye

Wilson

Humphrey

. Quantification of the mechanical behavior of carotid arteries from wild-type, dystrophin-deficient, and sarcoglycan-delta knockout mice. J Biomech. 2008;41:3213–3218. doi:10.1016/j.jbiomech.2008.08.012.

26.

Bersi

Ferruzzi

Eberth

Gleason

Jr Humphrey

. Consistent biomechanical phenotyping of common carotid arteries from seven genetic, pharmacological, and surgical mouse models. Ann Biomed Eng. 2014;42:1207–1223. doi:10.1007/s10439-014-0988-6.

27.

Sokolis

Sassani

Kritharis

Tsangaris

. Differential histomechanical response of carotid artery in relation to species and region: Mathematical description accounting for elastin and collagen anisotropy. Med Biol Eng Comput. 2011;49:867–879. doi:10.1007/s11517-011-0784-5.

28.

Sokolis

. Experimental investigation and constitutive modeling of the 3d histomechanical properties of vein tissue. Biomech Model Mechanobiol. 2013;12:431–451. doi:10.1007/s10237-012-0410-y.

29.

Sokolis

. Structurally-motivated characterization of the passive pseudo-elastic response of esophagus and its layers. Comput Biol Med. 2013;43:1273–1285. doi:10.1016/j.compbiomed.2013.06.009.

30.

Sokolis

Sassani

. Microstructure-based constitutive modelling for the large intestine validated by histological observations. J Mech Behav Biomed Mater. 2013;21:149–166. doi:10.1016/j.jmbbm.2013.02.016.

31.

Sokolis

. Identification and characterization of regional variations in the material properties of ureter according to microstructure. Comput Methods Biomech Biomed Engin. 2014;17:1653–1670. doi:10.1080/10255842.2012.761692.

32.

Berry

Greenwald

Rivett

. Static mechanical properties of the developing and mature rat aorta. Cardiovasc Res. 1975;9:669–678. doi:10.1093/cvr/9.5.669.

33.

Badreck-Amoudi

Patel

Kane

TPC

Greenwald

. The effect of age on residual strain in the rat aorta. J Biomech Eng. 1996;118:440–444. doi:10.1115/1.2796029.

34.

Tsamis

Rachev

Stergiopulos

. A constituent-based model of age-related changes in conduit arteries. Am J Physiol Heart Circ Physiol. 2011;301:H1286–H1301. doi:10.1152/ajpheart.00570.2010.

35.

Jee

WSS

Yao

. Overview: Animal models of osteopenia and osteoporosis. J Musculoskelet Neuronal Interact. 2001;1:193–207.

36.

Cox

Fischer

. Effects of sex hormones on the passive mechanical properties of rat carotid artery. Blood Vessels. 1978;15:266–276.

37.

Fischer

Swain

. Effect of sex hormones on blood pressure and vascular connective tissue in castrated and non-castrated male rats. Am J Physiol. 1977;237:617–621.

38.

Weizsäcker

Lambert

Pascale

. Analysis of the passive mechanical properties of rat carotid arteries. J Biomech. 1983;16:703–715. doi:10.1016/0021-9290(83)90080-5.

39.

Humphrey

Eberth

Dye

Gleason

. Fundamental role of axial stress in compensatory adaptations by arteries. J Biomech. 2009;42:1–8. doi:10.1016/j.jbiomech.2008.11.011.

40.

Fischer

. In vivo effects of estradiol on collagen and elastin dynamics in rat aorta. Endocrinology. 1972;91:1227–1232. doi:10.1210/endo-91-5-1227.

41.

Wolinsky

. Effects of estrogen treatment on the oophorectomized female rat aorta. Proc Soc Exp Biol Med. 1973;144:864–867. doi:10.3181/00379727-144-37699.

42.

Fischer

Swain

. Influence of contraceptive and other sex steroids on aortic collagen and elastin. Exp Mol Pathol. 1980;33:15–24. doi:10.1016/0014-4800(80)90003-9.

43.

Bal

Kurup

. Oestrogens and metabolism of glycosaminoglycans: Effect of oophorectomy and administration of oestradiol in female rats. Indian J Biochem Biophys. 1979;16:299–303.

44.

Natoli

Medley

Ahimastos

Drew

Thearle

Dilley

, et al. Sex steroids modulate human aortic smooth muscle cell matrix protein deposition and matrix metalloproteinase expression. Hypertension. 2005;46:1129–1134. doi:10.1161/01.HYP.0000187016.06549.96.

45.

Dobrin

Schwarcz

Baker

. Mechanisms of arterial and aneurysmal tortuosity. Surgery. 1988;104:568–571.

46.

Langille

Bendeck

Keeley

. Adaptations of carotid arteries of young and mature rabbits to reduced carotid blood flow. Am J Physiol Heart Circ Physiol. 1989;256:H931–H939.

47.

Learoyd

Taylor

. Alteration with age in the viscoelastic properties of human arterial wall. Circ Res. 1966;18:278–292. doi:10.1161/01.RES.18.3.278.

48.

Fung

. What are the residual stresses doing in our blood vessels? Ann Biomed Eng. 1991;19:237–249. doi:10.1007/BF02584301.

49.

Sokolis

. Effects of aneurysm on the directional, regional, and layer distribution of residual strains in ascending thoracic aorta. J Mech Behav Biomed Mater. 2015;46:229–243. doi:10.1016/j.jmbbm.2015.01.024.

50.

Prentice

. Postmenopausal hormone therapy and the risks of coronary heart disease, breast cancer, and stroke. Semin Reprod Med. 2014;32:419–425. doi:10.1055/s-0034-1384624.

51.

Aspray

Francis

. Treatment of osteoporosis in women intolerant of oral bisphosphonates. Maturitas. 2012;71:76–78. doi:10.1016/j.maturitas.2011.10.005.

52.

Kim

Choi

Ryu

. Is complementary and alternative therapy effective for women in the climacteric period? J Menopausal Med. 2015;21:28–35. doi:10.6118/jmm.2015.21.1.28.

53.

Mukwaya

Wong

Zhang

. Chinese herbal medicine for bone health. Pharm Biol. 2014;52:1223–1228. doi:10.3109/13880209.2014.884606.

54.

Simons

Gruppen

Bovee

Verbruggen

Vincken

. Prenylated isoflavonoids from plants as selective estrogen receptor modulators (phytoSERMs). Food Funct. 2012;3:810–827. doi:10.1039/c2fo10290k.

55.

Liu

Xia

. Chinese herbal medicines for treating osteoporosis. Cochrane Database Syst Rev. 2014;(3):CD005467.

56.

Fridez

Zulliger

Bobard

Montorzi

Miyazaki

Hayashi

Stergiopulos

. Geometrical, functional, and histomorphometric adaptation of rat carotid artery in induced hypertension. J Biomech. 2003;36:671–680. doi:10.1016/S0021-9290(02)00445-1.

57.

Tsamis

Stergiopulos

. Arterial remodeling in response to hypertension using a constituent-based model. Am J Physiol Heart Circ Physiol. 2007;293:H3130–H3139. doi:10.1152/ajpheart.00684.2007.

58.

Parker

Jacoby

Shoupe

Rocca

. Effect of bilateral oophorectomy on women’s long-term health. Womens Health (Lond Engl). 2009;5:565–576. doi:10.2217/whe.09.42.

59.

Nkonde-Price

Bender

. Menopause and the heart. Endocrinol Metab Clin North Am. 2015;44:559–564.