Using 4-vinylcyclohexene diepoxide as a model of menopause for cardiovascular disease

Abstract

There is a sharp rise in cardiovascular disease (CVD) risk and progression with the onset of menopause. The 4-vinylcyclohexene diepoxide (VCD) model of menopause recapitulates the natural, physiological transition through perimenopause to menopause. We hypothesized that menopausal female mice were more susceptible to CVD than pre- or perimenopausal females. Female mice were treated with VCD or vehicle for 20 consecutive days. Premenopausal, perimenopausal, and menopausal mice were administered angiotensin II (ANG II) or subjected to ischemia-reperfusion (I/R). Menopausal females were more susceptible to pathological ANG II-induced cardiac remodeling and cardiac injury from a myocardial infarction (MI), while perimenopausal, like premenopausal, females remained protected. Specifically, ANG II significantly elevated diastolic (130.9 ± 6.0 vs. 114.7 ± 6.2 mmHg) and systolic (156.9 ± 4.8 vs. 141.7 ± 5.0 mmHg) blood pressure and normalized cardiac mass (15.9 ± 1.0 vs. 7.7 ± 1.5%) to a greater extent in menopausal females compared with controls, whereas perimenopausal females demonstrated a similar elevation of diastolic (93.7 ± 2.9 vs. 100.5 ± 4.1 mmHg) and systolic (155.9 ± 7.3 vs. 152.3 ± 6.5 mmHg) blood pressure and normalized cardiac mass (8.3 ± 2.1 vs. 7.5 ± 1.4%) compared with controls. Similarly, menopausal females demonstrated a threefold increase in fibrosis measured by Picrosirus red staining. Finally, hearts of menopausal females (41 ± 5%) showed larger infarct sizes following I/R injury than perimenopausal (18.0 ± 5.6%) and premenopausal (16.2 ± 3.3, 20.1 ± 4.8%) groups. Using the VCD model of menopause, we provide evidence that menopausal females were more susceptible to pathological cardiac remodeling. We suggest that the VCD model of menopause may be critical to better elucidate cellular and molecular mechanisms underlying the transition to CVD susceptibility in menopausal women.

NEW & NOTEWORTHY Before menopause, women are protected against cardiovascular disease (CVD) compared with age-matched men; this protection is gradually lost after menopause. We present the first evidence that demonstrates menopausal females are more susceptible to pathological cardiac remodeling while perimenopausal and cycling females are not. The VCD model permits appropriate examination of how increased susceptibility to the pathological process of cardiac remodeling accelerates from pre- to perimenopause to menopause.

INTRODUCTION

Menopause is a condition that every woman will experience, and over the last 25 years, more women than men have died from cardiovascular disease (CVD) (CDC; 2019). This increase in CVD mortality in women results from loss of CVD protection associated with the onset of menopause, when there is sharp rise in CVD morbidity and mortality in women compared with men (14, 38, 53). Life expectance of women has now reached 82 (almost 35 yr longer than at the turn of the 20th century) (6). Yet, women typically undergo menopause at 50–60 yr of age, which means that women spend up to 40% of their life in menopause (6). Therefore, menopause and associated CVD risk must be considered as distinct from an aging or senescent woman. Despite longstanding knowledge that premenopausal women are protected from CVD, our fundamental understanding regarding the shift in CVD risk with menopause remains inadequate and impedes our ability to develop sex-specific therapeutic strategies to combat menopausal susceptibility to CVD.

Estrogen, which is no longer produced by ovarian tissue after menopause, is a naturally occurring steroid hormone that is positioned to play a unique role in cardioprotection. Estrogen signaling is complex, and multiple molecular, genetic, and cellular mechanisms have been suggested to underlie its protection against CVD (32, 77). Investigations into the cardioprotective effect of estrogen are complicated by findings in human studies compared with rodents. Generally, rodents with surgical removal of ovaries (ovariectomy) benefit from estrogen replacement when subjected to different cardiac pathological stimuli (7, 12, 58, 63, 73). However, the prospective Women’s Health Initiative (WHI) and the Heart and Estrogen/Progestin Replacement Study (HERS I and II) show increased CVD and stroke risk with estrogen replacement in menopausal women (25, 54). Subsequent analyses found women receiving estrogen who had recently transitioned to menopause (age 50–59) trended toward a reduced CVD risk, unlike older women (age 70–79 yr) receiving estrogen (55). The implication is that early estrogen replacement, perhaps during perimenopause, is protective against CVD (54). It remains unknown whether the suggested benefit of estrogen in perimenopausal women will translate into a benefit or detriment if estrogen therapy is continued into menopause. Therefore, combinatorial approaches to study estrogen-dependent mechanisms using appropriate models are required.

The critical barrier impeding a better understanding of how the drop of estrogen during menopause increases CVD risk is the lack of appropriate rodent models to study menopause. Most studies have used ovariectomy (OVX) as a model of menopause. However, this technique poorly recapitulates the natural, physiological transition to menopause that 90% of women experience and critically lacks the transitional, or perimenopausal, period. Our innovative strategy to overcome this barrier uses an ovary-intact mouse model of menopause induced by the occupational chemical 4-vinylcyclohexene diepoxide (VCD) (27, 42). Short-term daily dosing with VCD selectively targets primordial follicles of the ovaries with no evidence of necrotic changes, accelerating the natural process of follicular atresia to induce gradual ovarian failure and loss of estrogen production. This model preserves the important “perimenopause” transitional period and androgen-secreting capacity of residual ovarian tissue, analogous to menopausal women (42, 51). At this point, several studies demonstrate that VCD directly targets primordial and primary follicles of the ovary with no toxicity to nonovarian tissue [see Brooks et al. (5) for review]. We have successfully validated this model demonstrating that during perimenopause, females are protected from ANG-II hypertension and adverse remodeling in the kidney; menopausal females experience exacerbated hypertension, indicative of worse clinical outcomes and hallmarking increased CVD susceptibility (49).

Hypertension and coronary artery disease have sex-specific characteristics that also change throughout a woman’s life cycle (70). For example, women are at greater risk of developing resistant hypertension than men, especially as they reach menopause (59). Importantly, the incidence of myocardial infarction (MI) rises in menopausal women, and women with uncontrolled hypertension are sevenfold more likely to die from an ischemic event (4). In fact, the median age of the first acute MI is 10 years later in women than in men (1). In this series of experiments, we hypothesized that menopausal female mice are more susceptible to CVD than pre- or perimenopausal female mice. We mimicked the two most common forms of CVD, hypertension and ischemic heart disease, during pre-, peri-, and menopausal states. We then analyzed cardiac remodeling by echocardiography, cardiac hypertrophy, fibrosis, ischemic damage, and cellular and molecular markers. Overall, we demonstrated using the VCD model of menopause that perimenopausal mice, like premenopausal female mice, were protected from pathological ANG II-induced cardiac remodeling and ischemic heart disease while menopausal female mice were not, similar to humans.

MATERIALS AND METHODS

Treatment with 4-Vinylcyclohexene Diepoxide

All experiments were performed using protocols that adhered to guidelines and approved by the Institutional Animal Care and Use Committee at the University of Arizona and to 2011 National Institutes of Health guidelines for care and use of laboratory animals. At 3 mo of age, C57BL/6 female mice were randomized to premenopause, perimenopause, or menopause. Mice were weighed and given daily intraperitoneal injections of VCD (V3630; Sigma-Aldrich, St. Louis, MO) at a dose of 160 mg/kg for 20 consecutive days. Mice treated with sesame oil (SO), the solvent used for VCD, served as respective, premenopausal controls for perimenopausal and menopausal cohorts. To account for the age discrepancy between perimenopause and menopause groups, a premenopause (Pre-A) group was paired with the perimenopause group and a separate premenopause (Pre-B) cohort was paired with the menopause group leading to four separate experimental groups (Table 1).

Table 1.

Experimental groups

	Day 34	Day 60
Sesame oil	Premenopause-A (Pre-A)	Premenopause-B (Pre-B)
VCD	Perimenopause (Peri)	Menopause (Meno)

VCD, 4-vinylcyclohexene diepoxide.

After the 20-day injection period, estrous cycles were monitored daily by vaginal cytology to determine when cycling ceased and therefore when ovarian failure occurred. Mice were considered acyclic after 15 consecutive days in persistent diestrus. Estrous cycles were monitored closely throughout the study to confirm estrous cycles in VCD-treated mice and entry into either perimenopause or menopause as previously detailed by our group. In the current study, ovarian failure (menopause) occurred by 60 days after the onset of VCD dosing, whereas perimenopause was validated by 34 days after the onset of VCD dosing. To confirm ovarian failure, ovaries were collected postmortem, trimmed of adipose tissue, and fixed in Bouin fixative at the time of harvest. Samples were then transferred to 70% ethanol and paraffin embedded, and serial sections (thickness, 4 to 5 μm) were prepared by the University of Arizona Histology Laboratory. Sections were mounted and stained with hematoxylin and eosin. In each mouse, follicles were counted and classified by a blinded observer as primordial, primary, secondary, or antral, as previously described (9, 47, 49).

Treatment with Angiotensin II

Under inhaled isoflurane anesthesia, osmotic minipumps (model 1002, Azlet) containing ANG II or saline (infusion rate of 800 ng·kg⁻¹·min⁻¹; A9525; Sigma-Aldrich, St. Louis, MO) were implanted subcutaneously in the interscapular space. In the perimenopause group, pumps were implanted on day 34, 14 days following the completion of 20 days VCD dosing. In the menopause group, pumps were implanted 60 days from the start of VCD dosing, after the onset of ovarian failure (47, 49). Age-matched premenopausal, perimenopausal, and menopausal females implanted with osmotic minipumps containing saline served as controls for ANG II.

Blood Pressure Monitoring

Blood pressure readings were taken by noninvasive tail-cuff machine (MC4000; Hatteras Instruments) 1 day before ANG-II pumps were inserted and again after 14 days of ANG-II infusion. To acclimate the mice, blood pressure was recorded for 3 consecutive days before each final time point.

Echocardiography

For a noninvasive examination of morphological and functional changes associated with persistent ANG-II administration, transthoracic echocardiography was performed using a Visual Sonics Vevo 770 high-resolution imaging system (Visual Sonics, Toronto, ON, Canada). All groups underwent echocardiographic analysis before ANG-II treatment, during ANG-II treatment and immediately before euthanasia. A 25-MHz transducer was used. The chests of the animals were shaved with a chemical hair remover. Anesthesia was maintained by 1% isoflurane with oxygen. Body temperature was maintained using a heated platform. Respiratory rates and electrocardiograms were monitored throughout the study. Two-dimensional M-mode echocardiographic images were obtained from the parasternal short-axis views at the level of the midventricles. Cardiac chamber dimensions and the left ventricular wall thickness were measured. Left ventricular (LV) posterior wall thickness (LVPW) and internal dimension (LVID) were measured from the M-mode images. Relative wall thickness (RWT) [(LVPW/LVID) × 2] was calculated from the M-mode measurements. Data were analyzed off-line using Vevo 770 analytic software. The data were obtained in triplicate and averaged.

Cardiac Myosin Heavy-Chain Isoform

Snap-frozen ventricular tissue samples were prepared for SDS-PAGE to determine relative cardiac myosin heavy chain (MHC), as previously described (47). Briefly, samples were homogenized (Next Advance, Averill Park, NY) and diluted in a sample buffer that contained 8 M urea, 2 M thiourea, 3% SDS (wt/vol), 75 mM DTT, 0.03% bromophenol blue, and 0.05 M Tris·HCl (pH 6.8). MHC isoforms were separated on a 6% acrylamide resolving gel (37.5:1 crosslinked with DATD; Bio-Rad, Hercules, CA) and a 2.95% stacking gel (5.6:1 crosslinked with DATD; Bio-Rad) by using a SE600 Hoefer gel system (Hoefer, Holliston, MA) at 16-mA constant current. The gels were silver stained (Silver Stain Plus kite, Bio-Rad 161-0449), scanned (EPSON V750 PRO), and analyzed to determine the ratio of α-MHC to β-MHC isoforms using densitometry (LabImage 1D L340 N3, Kaplean BioImaging System, Leipzig, Germany). To determine relative MHC content, six to seven dilutions of each sample were analyzed so that readings stayed in the linear densitometric range. From the linear relationships determined for each MHC isoform, the relative MHC content of each isoform was extrapolated. Soleus (containing predominately types-Iβ and -IIa MHC) and neonatal rat ventricular cardiomyocytes (containing both α- and β-MHC) were used as standards.

Surgical Procedures

After confirmation of either perimenopause (34 days from the start of VCD injection) or menopause (60 days after the start of VCD injection), mice were anesthetized with an intraperitoneal injection of 250 mg/kg tribromoethanol (Sigma), intubated, and ventilated with 0.5–2.0% isoflurane (Phoenix Pharmaceuticals). A single injection of buprenorphine-SR (Reckitt Benckiser Healthcare) at 1 mg/kg body wt was given before surgery. An ischemia-reperfusion (I/R) protocol was used in this study. A left anterior thoracotomy was performed to expose the heart and the left coronary artery (LCA) visualized. The LCA was occluded using 8-0 suture compressing a small piece of tubing (PE-10) to prevent vessel damage during occlusion. Blanching of the myocardium was used to confirm ligation and ensure similar infarct size. After 30 min of occlusion, the ligature was removed, and the animal was allowed to recover for a 3-day reperfusion period.

Real-Time PCR

Total RNA was isolated from the left ventricles of vehicle (SO)- and VCD-treated female mouse hearts harvested at the end of each ANG-II or sham regimen by using the High Pure miRNA Isolation kit (Roche, Indianapolis, IN) according to the manufacturer’s protocol. Total cDNA was generated by using the NCode miRNA First-Strand cDNA Synthesis Kit (Invitrogen, Carlsbad, CA). Real-time PCR was done according to the Universal ProbeLibrary Assay (Roche) using the LightCycler 480 system (Roche). Primers and locked nucleic acid probes from the Roche Universal Probe Library were designed and chosen using the ProbeFinder Assay Design software (Roche, http://qpcr.probefinder.com). Relative expression of brain natriuretic peptide (BNP) and atrial natriuretic peptide (ANP) were quantified using the 2^-ΔΔCp method, where Cp is the measured crossing point at which PCR amplification begins its exponential phase determined by the second derivative maximum of each amplification curve using Light cycler 480 analysis software v. 3.3. Reference genes used for normalization were β-actin and GAPDH.

Semiquantitative Determination of Fibrosis/Collagen Content

Hearts were processed, embedded in paraffin, sectioned, and stained with Picrosirius red according to standard protocols. In brief, the paraformaldehyde fixed tissue was prepared using a tissue processor and embedded in paraffin. The embedded tissue sectioned at 7-μm thickness. The specimens were dewaxed, rehydrated, and stained with Picrosirius red to detect collagen fibers. Collagen content was quantified using previously established methods with minor modifications (15, 29, 30, 71, 72). Rehydrated heart sections were placed on a movable, rotating stage and visualized using a 25× strain free objective under polarized light using a Zeiss universal microscope (Axio Imager M1, Zeiss, Oberkochen, Germany), equipped with polarizing filters and an analyzer positioned at 90° to the polarizer above the objective. Disarray and fibrosis appear as yellow/orange birefringence under polarizing light, whereas highly ordered sarcomeres appear green (15, 29, 30, 71, 72). Five random images were digitized and then analyzed by a semiautomated imaging analysis program (AxioVision, Zeiss). Collagen content was expressed as a percentage of the total image area. For each sample, the five images were averaged and used as the representative percent collagen content. Five independent samples were used for each experimental group. Nonuniform myocytes exhibited no birefringence and were assumed to be indicative of myocellular disarray.

Determination of Infarct Size and Immunohistochemistry

Following the 72-h reperfusion period, mice were anesthetized with an intraperitoneal injection of 250 mg/kg tribromoethanol (Sigma), and the LCA was reoccluded at the same original site. Evans Blue (EB) dye (1%, Sigma) was then perfused retrograde through the aorta in situ, staining the whole heart except the left coronary bed and outlining the ischemic zone or area at risk (AAR). The hearts were excised, transversely sectioned at 1-mm thickness, and incubated in 1% 2,3,5-triphenyltetrazoliumchloride stain (TTC; T8877, Sigma, St. Louis, MO). After incubation, infarcted myocardium stained white [area of necrosis (AON)], viable tissue within AAR stained red, and perfused tissue remained blue. AAR and AON were measured in five to six sections from the top and bottom views resulting in 10–12 calculations per heart using image analysis software (ImageJ). Infarct size was reported as a percentage of AON to AAR.

Data and Statistical Analysis

Results are presented as means ± SE. We performed two-way ANOVA followed by a Bonferroni post hoc test to compare mean values. The percent change in echocardiographic parameters with ANG II was determined by comparing LVID, LVPW, or RWT from each mouse after 14-day ANG-II administration to the mean parameter of saline (control) animals and expressed as a percentage. Because we were unable to perform paired analysis to determine the extent of cardiac growth, the percent change in cardiac mass with ANG II was determined by comparing the normalized [heart weight to body weight (HW/BW)] mass of the heart from each mouse after 14-day ANG-II administration to the mean normalized (HW/BW) cardiac mass of saline (control) animals and expressed as a percentage. In addition, two-way ANOVA followed by a Bonferroni post hoc test or Student’s t test was performed to compare differences between infarct size. P values < 0.05 were considered statistically significant.

RESULTS

Menopausal Mice Are More Susceptible to Hypertension

Hypertension is the number one risk factor for CVD in women, and both CVD and hypertension dramatically increase after menopause (75, 76). Using the VCD model of menopause combined with the infusion of the hypertensive agent ANG II, we demonstrated that menopausal mice were more susceptible to hypertension induced by ANG II compared with perimenopausal and premenopausal mice similar to our previous work (49). When premenopausal control (sesame oil) and VCD-treated perimenopausal mice were compared, there were no significant differences in diastolic (100.5 ± 4.1 and 93.7 ± 2.9 mmHg for premenopausal and perimenopausal mice, respectively) or systolic (123.8 ± 4.8 and 122.4 ± 3.5 mmHg for premenopausal and perimenopausal mice, respectively) blood pressure measured at baseline, as shown in Fig. 1A. Fourteen days of ANG-II infusion significantly increased systolic blood pressure (main effect of ANG II, P < 0.001) in both premenopausal (152.3 ± 6.5 mmHg) and perimenopausal (155.9 ± 7.3 mmHg) mice.

Fig. 1.

Menopause exacerbates ANG-II hypertension. A: diastolic (DBP; circles) and systolic (SBP; triangles) blood pressure measured at baseline and at day 14 in ANG II-infused (triangles) premenopausal (Pre-A; open symbols) or perimenopausal (Peri; closed symbols) female mice. Main effect of ANG-II infusion (P < 0.001) by 2-way ANOVA. B: DBP (circles) and SBP (triangles) measured at baseline and at day 14 in ANG II-infused (triangles) in premenopausal (Pre-B; open symbols) or menopausal (Meno; closed symbols) female mice. Main effect of ANG-II infusion (P < 0.001) with a significant interaction between menopause and ANG-II infusion (P < 0.0001) by 2-way ANOVA. *P < 0.01 from controls by post hoc analysis. Data are presented as means ± SE; n = 12–18 mice/group.

A separate cohort of VCD-treated mice were allowed to progress into ovarian failure (menopause), which was confirmed by vaginal cytology and ovarian histology as previously described (data not shown) (47, 49). Entering a state of menopause did not adversely affect blood pressure; diastolic blood pressure was similar between menopausal (95.2 ± 5.5 mmHg) and age-matched premenopausal (95.6 ± 4.3 mmHg). Systolic blood pressure showed a similar pattern between menopausal (121.2 ± 3.0 mmHg) and premenopausal mice (119.2 ± 1.9 mmHg) (Fig. 1B). Upon confirmation of menopause, premenopausal and menopausal mice were implanted with ANG II-releasing osmotic minipumps. Fourteen days of ANG-II infusion caused a significant elevation in diastolic and systolic blood pressure (main effect of ANG II, P < 0.0001) in premenopausal (diastolic, 114.7 ± 6.2 mmHg; systolic, 141.7 ± 5.0 mmHg) and menopausal (diastolic, 130.9 ± 6.0 mmHg; systolic, 156.9 ± 4.8 mmHg) mice. A significant interaction (P < 0.0001) occurred between ANG-II administration and menopause; post hoc analysis indicated that this interaction resulted from the significant (P < 0.01) elevation in diastolic and systolic blood pressure when comparing ANG II-treated menopausal and premenopausal mice (Fig. 1B). This increased susceptibility to ANG-II hypertension in menopausal mice confirms the data previously reported by our laboratory (49).

Menopausal Mice Are More Susceptible to Pathological Cardiac Hypertrophy

Echocardiography.

ANG II is a potent stimulator of pathological cardiac remodeling (11). Concomitant with exacerbation of menopausal hypertension over all other groups, menopausal mice subjected to ANG-II administration displayed worsening of pathological cardiac remodeling. To assess the impact of ANG II on in situ ventricular morphometry, high-resolution two-deminsional serial echocardiography was performed on premenopausal, perimenopausal, and menopausal animals with and without 14 days of ANG II. Several lines of evidence indicate that ANG II instigated cardiac hypertrophy in all experimental groups, which was worsened by menopause. The functional and morphometric data from all experimental groups are summarized in Table 2 and displayed in Fig. 2.

Table 2.

Cardiac hypertrophy and echocardiographic measurements

Study	Pre-A	Peri	Pre-A-ANG II	Peri-ANG II
Perimenopause
n	18	16	15	12
EF, %	50.8 ± 2.7	50.0 ± 2.7	54.2 ± 1.5	52.1 ± 3.3
LVIDd, mm	3.65 ± 0.06	3.78 ± 0.11	3.49 ± 0.11*,†	3.52 ± 0.10*,†
LVPWd, mm	0.64 ± 0.02	0.64 ± 0.02	0.77 ± 0.06*,†	0.74 ± 0.03*,†
RWTd	0.38 ± 0.02	0.38 ± 0.01	0.45 ± 0.05*,†	0.43 ± 0.03*,†
BW, g	20.2 ± 0.3	19.0 ± 0.3	20.0 ± 0.4	18.2 ± 0.4
HW, mg	91.2 ± 1.6	92.5 ± 2.0	99.04 ± 0.2*	98.3 ± 2.5
HW/BW, mg/g	4.68 ± 0.07	4.81 ± 0.08	5.06 ± 0.08*	5.24 ± 0.18*,†
HW/TL, mg/mm	5.10 ± 0.07	4.99 ± 0.12	5.51 ± 0.09†	5.49 ± 0.15†

	Pre-B	Meno	Pre-B-ANG II	Meno-ANG II
Menopause
n	14	14	15	15
EF, %	51.5 ± 3.2	51.1 ± 2.7	59.5 ± 2.6	58.3 ± 4.2
LVIDd, mm	3.71 ± 0.10	3.69 ± 0.05	3.26 ± 0.10‡,§	3.36 ± 0.08‡,§
LVPWd, mm	0.67 ± 0.01	0.64 ± 0.02	0.89 ± 0.06‡,§	0.87 ± 0.06‡,§
RWTd	0.35 ± 0.01	0.38 ± 0.01	0.51 ± 0.04‡,§	0.49 ± 0.02‡,§
BW, g	22.0 ± 0.6	23.5 ± 0.4	22.6 ± 0.4	21.6 ± 0.4
HW, mg	100.3 ± 3.0	102.3 ± 2.2	112.4 ± 3.7‡	112.5 ± 1.8‡
HW/BW, mg/g	4.48 ± 0.13	4.31 ± 0.11	4.82 ± 0.13§	5.13 ± 0.06‡,§
HW/TL, mg/mm	5.46 ± 0.14	5.57 ± 0.12	5.89 ± 0.19	6.14 ± 0.09‡,§

Values are means ± SE. Measurements are shown for perimenopausal [premenopausal (Pre-A), perimenopausal (Peri), Pre-A + ANG II, and Peri + ANG II] and menopausal [premenopausal (Pre-B), menopausal (Meno), Pre-B + ANG II, and Meno + ANG II] female mice. EF, ejection fraction; LVIDd, left ventricular (LV) internal diameter at end diastole; LVPWd, LV posterior wall thickness at end diastole; RWTd, relative wall thickness at end diastole; BW, body weight (in g); HW, heart weight (in mg); HW/BW, heart weight normalized to body weight (in mg/g); HW/TL, heart weight normalized to tibia length (in mg/mm). Significance by post hoc analysis as follows:

^*P < 0.01 from Pre-A;

^†P < 0.01 from Peri;

^‡P < 0.01 from Pre-B;

^§P < 0.01 from Meno (non-ANG II treated).

Fig. 2.

Effect of ANG-II infusion on echocardiographic parameters of ventricular function and morphometry in pre- (Pre-A; Pre-B), peri- (Peri), and menopausal (Meno) female mice. A–C: control and perimenopausal female mice. D–F: control and menopausal female mice. LVIDd, left ventricular (LV) internal diameter at end diastole; LVPWd, LV posterior wall thickness at end diastole; RWTd, relative wall thickness at end diastole calculated as 2·LVPWd/LVIDd and represents a measure of LV eccentricity/concentricity. Main effect of ANG-II infusion (P < 0.001) by 2-way ANOVA within Peri or Meno groups. *P < 0.01 from controls by post hoc analysis. Data are presented as means ± SE; n = 12–18 mice/group.

A two-way ANOVA demonstrated a significant (P < 0.001) impact of ANG II on left ventricular internal dimension during diastole (LVIDd) and left ventricular posterior wall during diastole (LVPWd) (Fig. 2, A, B, D, and E) in all groups. In the perimenopausal group, post hoc analysis demonstrated that the impact of ANG II resulted in a significant decrease in LVIDd (shrinking of LV chamber size; P < 0.01) in pre- (−4.5 ± 3.1%) and perimenopausal (−6.6 ± 2.6%) following ANG-II administration. Similarly, there was a significant increase in LVPWd (thickening of LV wall thickness; P < 0.01) following ANG-II administration in pre- (20.3 ± 9.4%) and perimenopausal (15.6 ± 5.2%). In the menopausal group, ANG II also resulted in a significant (P < 0.01) decrease by post hoc analysis in LVIDd in pre- (−12.1 ± 2.7%) and menopausal (−9.0 ± 2.3%). Similarly, there was a significant (P < 0.01) increase in LVPWd following ANG-II administration in pre- (33.4 ± 9.4%) and menopausal (36.6 ± 8.7%). The effect of ANG II on LVPWd and LVIDd was similar in premenopausal, perimenopausal, and menopausal groups (Table 2). The directional impact of ANG II on LVIDd and LVPWd are clear indicators of ventricular hypertrophy.

The geometric relationship between LVIDd and LVPWd, or relative wall thickness during diastole (RWTd), allows us to determine both the extent and type of cardiac hypertrophy, i.e., concentric versus eccentric cardiac hypertrophy (23, 35, 65). We found a significant effect of ANG-II administration on RWTd (P < 0.001), which we could attribute as a significant increase in RWTd (Fig. 2, C and F; P < 0.01 in both Pre-A (16.7 ± 13.1%), Pre-B (45.2 ± 11.0%), perimenopausal (12.6 ± 7.9%) and menopausal (27.6 ± 4.8%) groups. Again, post hoc analysis validated a significant impact of ANG II on RWTd over control counterparts. Because an increase in RWTd is characterized by a disproportionate increase in wall thickness relative to chamber size, we can categorize the ANG-II effect on the heart as concentric cardiac hypertrophy. Again, the effect of ANG II on RWTd was similar in premenopausal, perimenopausal, and menopausal groups (Table 2).

Cardiac hypertrophy.

At the conclusion of the 14-day ANG-II infusion protocol, hearts were excised and weighed to determine the extent of cardiac hypertrophy. As illustrated in Fig. 3 and Table 2, ANG II had a significant effect on heart weight in all experimental groups, both in comparing absolute values and when normalized to body weight or tibial length. When comparing normalized heart weights as heart weight/body weight (HW/BW) between ANG II-treated and control counterparts, ANG-II infusion in menopausal females elicited a greater (P < 0.01 by post hoc analysis) relative increase in cardiac mass compared with perimenopausal mice (Fig. 3C) (Pre-A, 7.5 ± 1.4%; Peri, 8.3 ± 2.1%; Pre-B, 7.7 ± 1.5%; and Meno, 15.9 ± 1.0%). Although ANG II significantly decreased LVIDd while decreasing LVPWd, RWTd, and HW/BW for pre-, peri-, and menopausal groups, an elevated RWTd coupled with increased cardiac mass only occurred in the menopausal group and is associated with elevated mortality in humans (23, 35).

Fig. 3.

Cardiac adaptation to ANG-II infusion in pre- (Pre-A; Pre-B), peri- (Peri), and menopausal (Meno) female mice. A and B: cardiac adaptation was determined by normalizing heart weight to body weight (HW/BW; mg/g). ANG-II infusion significantly increased normalized heart weight in Pre-A and Peri (A) and Pre-B and Meno (B) female mice compared with controls. C: ANG II induced a significantly greater increase in HW/BW (%change from control) in menopausal mice compared with premenopausal or perimenopausal mice. Main effect of ANG-II infusion (P < 0.01) by 2-way ANOVA. *P < 0.01 from control and Peri hearts by post hoc analysis. Data are presented as means ± SE; n = 12–18 mice/group.

Menopausal Mice Demonstrate Cellular Pathological Cardiac Remodeling

mRNA expression.

There is an array of pathological markers that are characteristic of CVD, including transcriptional regulators, expression of hypertrophic genes and associated proteins, myocyte ultrastructure, and fibrosis (20, 45, 67). Considering the exacerbation of cardiac hypertrophy in menopausal females, we hypothesized that menopausal females will similarly display an exacerbation of pathological gene markers. Accordingly, we performed real-time quantitative PCR for brain natriuretic peptide (BNP) and atrial natriuretic peptide (ANP), both hallmarks of pathological cardiac stress (66). When BNP levels in the heart were measured, there was a significant impact of VCD treatment (main effect of VCD; P < 0.05 and P < 0.01 for perimenopause and menopause, respectively) in both the perimenopause and menopause groups (Fig. 4, A and B). A main effect of VCD suggests that the transition to menopause, which includes perimenopause may alter BNP expression. In the menopausal group, there was a main effect of ANG II (P < 0.01) that resulted in a significant decrease of BNP levels compared with control groups by post hoc analysis (P < 0.002 vs. all other groups).

Fig. 4.

Effect of ANG-II infusion on cardiac expression of brain natriuretic peptide (BNP) and atrial natriuretic peptide (ANP) in pre- (Pre-A; Pre-B), peri- (Peri), and menopausal (Meno) female mice. Quantitative real-time PCR was used to determine relative expression of BNP and ANP in left ventricular tissue from pre-, peri-, and menopausal mice. A: scatter plot summary of BNP mRNA expression in Peri female hearts compared with Pre-A hearts. Main effect of perimenopause (P < 0.01) by 2-way ANOVA. B: scatter plot summary of BNP mRNA expression in Meno female hearts compared with Pre-B hearts. Main effect of menopause and ANG-II infusion (P < 0.01) by 2-way ANOVA. *P < 0.01, from all other groups by post hoc analysis. C: scatter plot summary of ANP mRNA expression in Peri female hearts compared with Pre-A hearts. Main effect of ANG-II infusion (P < 0.01) by 2-way ANOVA. D: scatter plot summary of ANP mRNA expression in Meno female hearts compared with Pre-B hearts. Main effect of ANG-II infusion with a significant interaction (P < 0.01) by 2-way ANOVA. *P < 0.01 from controls and Meno-ANG II by post hoc analysis. Data are presented as means ± SE; n = 4–6 for each group.

ANG-II infusion increased ANP mRNA expression in premenopausal and perimenopausal hearts as a main effect (P < 0.01) and by post hoc analysis (P < 0.05 compared with controls; Fig. 4C). On the other hand, ANG II did not increase ANP levels in menopausal hearts, resulting in a significant interaction between VCD treatment and ANG II (P < 0.01; Fig. 4D).

Myosin heavy chain expression.

A recapitulation of β-myosin heavy chain (β-MHC) protein is a hallmark of pathological cardiac remodeling, and we have previously shown that VCD treatment alone does not induce β-MHC expression (47). Because ANG II is potent stimulator of β-MHC expression (40, 78), we measured the amount of β-MHC expression relative to α-MHC expression in all experimental groups. Nonpathological levels of β-MHC protein were present in the perimenopausal group (Pre-A, 1.9 ± 1.4%; Peri, 2.4 ± 1.3%) and menopausal group (Pre-B, 2.8 ± 1.4%; Meno, 2.4 ± 0.6%) and were consistent with previous studies (Fig. 5) (47, 60). ANG-II infusion significantly (main effect of ANG-II infusion P < 0.001; P < 0.01 from controls by post hoc analysis) elevated β-MHC expression relative to α-MHC expression in the perimenopausal group (Pre-A, 5.5 ± 1.4%; Peri, 6.7 ± 2.7%) and menopausal group (Pre-B, 6.5 ± 1.4%; Meno, 8.7 ± 2.4%; Fig. 5). These data indicate that the exacerbation of cardiac hypertrophy in menopausal females is not secondary to β-MHC expression.

Fig. 5.

Effect of ANG-II infusion on α-myosin heavy chain (α-MHC) and β-MHC expression in pre- (Pre-A; Pre-B), peri- (Peri), and menopausal (Meno) mice. Top: representative SDS-PAGE gel following silver staining shows separation of α-MHC and β-MHC in 6 to 7 protein concentrations (from 0.25×–20×) of pre-, peri-, and menopausal hearts with or without ANG II; soleus muscle was used as β-MHC control (see materials and methods). Bottom: scatter plot summary of β-MHC relative to α-MHC protein expression in Pre-A compared with Peri (A) or Pre-B compared with Meno (B) with and without ANG II. Main effect of ANG-II infusion by 2-way ANOVA (P < 0.001). *P < 0.01 from control groups by post hoc analysis. Data are presented as means ± SE; n = 4–6 for each group.

Collagen deposition.

Following sustained ANG-II infusion, fibroblasts differentiate into collagen-producing myofibroblasts, leading to collagen reorganization and accumulation (43, 48, 74). Therefore, fibrosis and myocellular disarray were evaluated in excised and fixed hearts by Picrosirius red staining (34, 47). Quantitative analysis for Picrosirius red under polarized light indicated that the induction of perimenopause (0.30 ± 0.15%) or menopause (0.59 ± 0.10%) did not impact myocardial fibrosis compared with premenopausal control (Pre-A 0.44 ± 0.09%; Pre-B 0.36 ± 0.08%) counterparts (Fig. 6). Following 14 days of ANG-II infusion, there was an increase in collagen deposition/fibrosis, resulting in a significant main effect of ANG-II treatment (P < 0.01) in the hearts of pre- (1.11 ± 0.26%) and perimenopausal (0.91 ± 0.34%) mice. It should be noted that the increase in mean differences was not significant by post hoc analysis, indicating that the “collective” increase in collagen deposition following ANG II, not a single experimental group, underlays the significant main effect of ANG II in the perimenopausal group. While the degree of fibrosis in perimenopausal hearts was indistinguishable from premenopausal hearts after ANG II, ANG II induced a threefold increase in collagen deposition in menopausal (2.81 ± 0.22%) hearts compared with premenopausal (1.09 ± 0.11%) hearts (main effect of ANG II and menopause with a significant interaction between ANG II and menopause; P < 0.0001 for all 3 parameters). Post hoc analysis validated that collagen deposition in ANG II-treated menopausal hearts was significantly elevated over all other groups (P < 0.001).

Fig. 6.

ANG II-induced collagen deposition in the hearts of pre- (Pre-A; Pre-B), peri- (Peri), and menopausal (Meno) female mice. Myocardial sections were stained with Picrosirius red to analyze collagen deposition. Thin collagen fibers have green birefringence and thick collagen fibers have bright yellow/orange birefringence under polarization light microscopy. Top: polarized microscopic images demonstrate enhanced myocardial collagen deposition (fibrosis) in all ANG II-infused experimental groups. Bottom: birefringence of collagen fibers was quantified using a semiautomated imaging analysis program and are expressed as percent fibrosis (collagen area) of total analyzed area. A: scatter plot summary of collagen deposition in Pre-A compared with Peri hearts with and without ANG II. Main effect of ANG-II infusion by 2-way ANOVA (P < 0.01). B: scatter plot summary of collagen deposition in Pre-B compared with Meno hearts with and without ANG II. Main effect of ANG-II infusion (P < 0.0001) and menopause (P < 0.0001) with a significant interaction between ANG II and menopause (P < 0.0001) by 2-way ANOVA. *P < 0.01 from premenopausal controls; #P < 0.001 from all other groups by post hoc analysis. Data are presented as means ± SE; n = 6 for each group.

Myocardial Infarction by I/R Is Exacerbated by Menopause

Because we contend that the use of VCD-induced menopause provides a physiologically appropriate model, we asked whether VCD-induced menopausal female mice experienced an exacerbation of cardiac injury from MI. Accordingly, we created an MI by I/R in perimenopausal and menopausal mice. We present the first evidence that hearts of menopausal female mice were more sensitive to cardiac injury from MI than perimenopausal or premenopausal females. We quantified myocardial damage by percent infarct size defined as the percentage of AON relative to AAR (Fig. 7). On average, menopausal females displayed a significant (P < 0.01) increase in infarct area (40.9 ± 5.5%) when compared with perimenopausal (18.0 ± 5.6%) and premenopausal (Pre-A, 16.2 ± 3.3%; Pre-B, 20.1 ± 4.8%) groups. Although previous work has demonstrated loss of protection from MI following surgical removal of the ovaries (7, 39, 81), these data represent the first evidence that menopausal females were more susceptible, or hypersensitive, to cardiac injury by MI. After the determination of infarct size, sections were fixed in 10% formalin overnight, paraffin embedded, and sectioned for staining with hematoxylin and eosin (H&E) and Picrosirius red. Using H&E and Picrosirius red, we validated EB and TTC staining; H&E and Picrosirius red staining of infarcted hearts matched infarct location and size with the EB and TTC staining (Fig. 7).

Fig. 7.

Extent of cardiac necrosis after ischemia-reperfusion in pre- (Pre-A; Pre-B), peri- (Peri), and menopausal (Meno) female mice. Left: Evans Blue dye (EBD)/2,3,5-triphenyltetrazoliumchloride (TTC) staining and hematoxylin and eosin (H&E) staining of a cross section of heart tissue from pre-, peri-, and menopausal female mice after ischemia-reperfusion. Black trace, area at risk (AAR); white trace, area of necrosis (AON). Right: scatter plot representation of the percent infarct size determined as AON-to-AAR ratio × 100 in the hearts of pre-, peri-, and menopausal mice. *P < 0.01 from Pre-B group; n = 9 for each group.

DISCUSSION

CVD remains the leading cause of death in both men and women. Yet, only 54% of women are aware of their CVD risk despite concerted efforts to educate women about CVD (44). It is now appreciated that estrogen loss due to the natural transition to menopause masks or underlies cardiac symptoms that are atypical compared with men (10, 79). Therefore, many women, particularly menopausal women, are unable to link their symptoms to CVD. One obstacle that has stalled clinical translation of studies into menopausal susceptibility to CVD is the lack of appropriate, preclinical rodent models mirroring progressive ovarian failure, i.e., one that includes a transition from perimenopause into menopause. Most studies have used surgical removal of ovaries (ovariectomy) as a model of menopause. However, this technique poorly recapitulates the natural physiological transition to menopause that 90% of women experience. Our innovative strategy to overcome these limitations takes advantage of an ovary-intact mouse model of menopause, using the occupational chemical VCD (27, 42). Short-term daily dosing with VCD selectively targets primordial follicles of the ovaries, accelerating the natural process of follicular atresia, and induces gradual ovarian failure. This model preserves the important “perimenopause” transitional period and androgen-secreting capacity of residual ovarian tissue, analogous to menopausal women (42, 51).

Our finding that infusion of ANG II at a pressor dose (hypertension causing) instigated a greater increase in blood pressure validated our previous work in this model (49). For this study, we chose a pressor dose of ANG II to mimic the clinical setting of “uncontrolled hypertension” in women, especially menopausal women. Despite the widespread clinical use of drugs that induce renin-angiotensin aldosterone system (RAAS) blockade (angiotensin-converting enzyme inhibition or angiotensin receptor blocker), a significant number of women are intolerant to RAS blockade or do not have well-controlled blood pressure (19, 69). Consequently, aberrant cardiac remodeling in response to excessive RAAS system and/or hypertension remains highly and clinically relevant.

Cardiac hypertrophy is a powerful predictor of adverse CVD outcomes (26, 28, 56, 80). Using high-resolution echocardiography, we determined in situ LV chamber dimensions. In general, mice treated with ANG II demonstrated patterns of hypertrophy consistent with the clinical literature. Using the geometric relationship between ventricular chamber size and ventricular wall thickness, or RWTd, we determined both the extent and type of cardiac hypertrophy, i.e., concentric versus eccentric cardiac hypertrophy (23, 35, 65). ANG II significantly increased RWTd for both premenopausal and menopausal groups to a similar extent. However, excised heart weights indicated that menopausal mice had greater hypertrophy. An elevated RWTd coupled with increased cardiac mass (concentric hypertrophy), as occurred in only the menopausal group, is associated with elevated mortality (23, 35). Moreover, patients with established concentric hypertrophy due to hypertension have the most severe clinical outcomes (22, 56). The suggestion is that menopausal females are more susceptible to more adverse cardiac pathology and an increased mortality risk.

Our laboratory and others have characterized an array of pathological markers of CVD, including transcriptional regulators, expression of hypertrophic genes and associated proteins, myocyte ultrastructure and fibrosis, cardiac function, and apoptosis (20, 34, 45, 67). Hallmarks of pathological cardiac stress include expression of ANP and BNP (66). The lack of ANG-II effect on BNP expression in pre- and perimenopausal mice is not surprising considering that women may be protected against CVD-induced increases in BNP when estrogen is present (21, 41). Interestingly, a recent cross-sectional study of menopausal women demonstrates a decrease in BNP with each 1-yr increase in menopause age. This finding may underlie our observation that BNP decreases with ANG-II treatment in menopausal females (17).

Not surprisingly, ANP was elevated in response to ANG II in both premenopausal and perimenopausal mice. However, ANG II did not increase ANP expression in menopausal mice. This may result from a reduced capacity to release ANP after estradiol depletion (36). A recapitulation of β-MHC protein is a hallmark of pathological cardiac remodeling. We have previously shown that VCD treatment alone does not induce β-MHC expression (47). In the present study, ANG II-induced increases in β-MHC protein were equivalent in hearts from premenopausal and menopausal mice. These data indicate that the exacerbation of cardiac pathology in menopausal females is not secondary to enhanced upregulation of β-MHC expression and suggest other factors mediate pathological responses to ANG II after menopause.

Previous work exhaustively validates that systemic or local ANG-II production induces interstitial and perivascular fibrosis and cardiac hypertrophy (3, 50). Progression to heart failure is driven by pathological remodeling including worsening fibrosis. Enhanced susceptibility to cardiac hypertrophy and concentric remodeling in menopausal mice from the present study was also accompanied by increased cardiac fibrosis, another important indicator of pathological cardiac remodeling, compared with all other groups (34).

We also show that menopausal female mice were more susceptible to cardiac injury following an I/R protocol, compared with perimenopausal or premenopausal mice. Previous studies where estrogen loss was achieved by surgical removal of ovaries as the model for menopause show similar increases in cardiac damage after an ischemic event (8, 61). However, to our knowledge, this is the first study to establish that mice retain similar protection against cardiac injury during perimenopause and in the context of a more physiological, natural transition to menopause. Menopause contributes to sex-specific risks such as increased susceptibility to hypertension, dyslipidemia, small-plaque erosion or rupture, and microvascular dysfunction (46). This supports the conclusion that the complex response orchestrated by estrogen, involved in post-cardiac injury phenotype recovery in females, is lost as females naturally transition into menopause. This loss of estrogen signaling, among other factors, contributes to a worse overall response to I/R injury.

Potential Clinical Value

Our group and others have concentrated significant effort toward identifying sex dimorphisms in cardiac adaptation to a variety of stimuli including hypertension, ischemic heart disease, hypertrophic cardiomyopathies FHC, and exercise (24, 31, 33, 57, 60, 62, 64). While the scientific and clinical communities acknowledge sex disparities in CVD, the mechanisms responsible for worsening cardiac pathology in menopausal women are unknown (2, 13, 38, 53). With the development of the VCD model of menopause, we are now able to overcome a critical barrier that has stalled clinical translation of studies into menopausal susceptibility to CVD. For example, the WHI study reported that only a small percentage of hypertensive women had their blood pressure under control (36.1%, overall average), with responsiveness worsening with increasing age (69). Perhaps more importantly is the observation that up to 20% of women display signs of intolerance (cough, hypotension, angioneurotic edema) to ACE inhibitors, a standard hypertension therapy (19).

Another key sex-specific characteristic is that almost two-thirds (64%) of women who suddenly die of coronary heart disease have no previous symptoms or symptoms dissimilar to men (52). We stress that pathological cardiac remodeling as a result of CVD in women lies on a continuum and will ultimately require more sophisticated approaches, like the VCD model, to delineate these limitations. For example, the VCD mouse model has the experimental flexibility to instigate menopause at any time point and to control the length of the perimenopausal transition depending on the time and extent of VCD dosing. To illustrate the significance of this concept, we report limited pathological consequences to CVD during perimenopause, which contradicts a recent study that found functional dysregulation in perimenopausal hearts (18). The differences between the two studies can be potentially explained by a shorter VCD dosing period (15 days) compared with 20 days of VCD dosing employed in our study protocol. Shortening the dosing period to 15 days lengthens the perimenopausal transition and presumably the underlying cardiac remodeling. In addition, the aforementioned study (18) used a different mouse strain (CD1) from our study (C57Bl6), which is known to impact cardiac remodeling (37, 68). Still, a direct comparison between the two studies is challenged by these differences in dosing and strain. Nevertheless, future studies can be carefully designed to control for unique strain differences and dosing protocols to better elucidate cellular and molecular mechanisms underlying our observations.

Moreover, the VCD model permits uncoupling reproductive senescence from physiological aging. In the current study, we initiated the VCD dosing protocol at 3 mo of age when mice are sexually mature but may be considered adolescent (16). Therefore, the transition to menopause occurs when the mouse can be considered mature yet far from middle aged or old. Reproductive senescence in rodents is complex and very often inconsistent with humans. However, experimental flexibility is critical considering that women do not enter the peri- to menopause transition at the same time and experience irregular cycling for different periods of length (5). Collectively, studies using the VCD model of menopause have transformational potential and will provide fundamental biology and new knowledge defining menopausal hypersensitivity to CVD. We believe that our data and models will assist many other researchers by providing a general view coupled with some mechanistic insight to more quickly advance this underrepresented area of research.

GRANTS

This work was supported by National Institutes of Health Grant R01-HL-098256, National Mentored Research Science Development Award K01-AR-052840, and Independent Scientist Award K02-HL-105799; American Heart Association Grant 16GRNT31390006 (to J.P.K.), Computational and Mathematical Modeling of Biomedical Systems Interdisciplinary Training Grant GM-004905, and Cardiovascular Sciences Interdisciplinary Training Grants HL-007249 and HL-00795515; and the Sarver Heart Center at the University of Arizona and the Steven M. Gootter Foundation.

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

J.P.K., J.N.S., J.A.R., E.C., L.A.M., H.C., and Y.L. conceived and designed research; J.P.K., J.N.S., J.A.R., E.C., L.A.M., H.C., and Y.L. performed experiments; J.P.K., J.N.S., J.A.R., E.C., L.A.M., H.C., and Y.L. analyzed data; J.P.K., J.N.S., J.A.R., E.C., L.A.M., H.C., and Y.L. interpreted results of experiments; J.P.K., J.N.S., J.A.R., E.C., M.A.L.-P., L.A.M., and Y.L. prepared figures; J.P.K., M.A.L.-P., D.K.C., R.S., D.P.P., and H.L.B. drafted manuscript; J.P.K., M.A.L.-P., D.K.C., R.S., D.P.P., and H.L.B. edited and revised manuscript; J.P.K., J.N.S., J.A.R., E.C., M.A.L.-P., D.K.C., R.S., L.A.M., H.C., Y.L., D.P.P., and H.L.B. approved final version of manuscript.