Excessive Hypercholesterolemia in Pregnancy Impairs Later‐Life Maternal Vascular Function in Rats

Abstract

Background

Preeclampsia is a risk factor for the development of later‐life cardiovascular disease. However, the underlying mechanisms are poorly understood. Excessive hypercholesteremia in pregnancy induces a preeclampsia‐like phenotype, but whether this also impacts maternal vascular function later in life has not been fully characterized.

Methods and Results

Sprague Dawley rats received a control diet (CD) or a high‐cholesterol (HCD) diet from gestational day 6 to 20, after which maternal vascular function was assessed 3 months postpartum. Exposure to an HCD in pregnancy reduced later‐life endothelium‐dependent vasodilation in carotid arteries (−15.24±3.27%), which was mediated via prostaglandin H synthase 2. There were no differences in vasodilation between CD and HCD postpartum rats in the mesenteric arteries, coronary arteries, or aortas. Vasoconstriction to phenylephrine increased in carotid arteries (61.02±21.48%) and reduced in aortas (−23.24±6.19%) of the HCD postpartum group versus CD dams, without differences in mesenteric and coronary arteries. The increased vasoconstriction in carotid arteries was due to lower nitric oxide modulation of constriction. Moreover, carotid artery myogenic response was reduced (−37.68±10.07%) and stiffness was increased (19.67±6.21%) in the HCD postpartum rats compared with CD along with decreased elastin density (−20.85±4.52%). The impact of the HCD on vascular function did not occur in age‐matched never‐pregnant female rats.

Conclusions

Excessive hypercholesterolemia in pregnancy impairs later‐life maternal vascular function in rats with varying impacts across different vascular beds. Understanding mechanisms for pregnancy‐specific excessive hypercholesterolemia provides avenues for targeted intervention strategies to reduce the burden of cardiovascular disease in women who had a complicated pregnancy.

Nonstandard Abbreviations and Acronyms

CCRC

cumulative‐concentration response curve

CD

control diet

E_max

maximum response

eNOS

endothelial NO synthase

GD

gestational day

HCD

high‐cholesterol diet

L‐NAME

N(G)‐nitro‐l‐arginine methyl ester hydrochloride

pEC₅₀

negative log of the drug concentration that produces 50% of the maximum response

PGHS

prostaglandin H synthase

PGHS2

prostaglandin H synthase isoform 2

PSS

physiological salt solution

Clinical Perspective.

What Is New?

• Exposure to a high‐cholesterol diet during pregnancy in rats impaired maternal vascular function 3 months postpartum, with varying impacts across different vascular beds.

• The conduit arteries (ie, carotid arteries and aortas), rather than the resistance‐sized vessels (ie, mesenteric and coronary arteries), were primarily impacted later in life after being exposed to excessively high cholesterol levels during pregnancy.

• The impact of the high‐cholesterol diet on vascular function was specific to pregnancy and did not occur in the age‐matched never‐pregnant female rats.

What Are the Clinical Implications?

• Excessive hypercholesterolemia in pregnancy predisposes the mother to the development of preeclampsia; moreover, a history of preeclampsia is a recognized risk factor for the development of cardiovascular disease in women later in life.

• For this reason, the maternal lipid profile may be a useful test during pregnancy to identify women at risk for preeclampsia, as well as those who may be at an increased risk of developing cardiovascular disease later in life.

Preeclampsia, a significant cause of maternal–fetal morbidity and death, is also a recognized risk factor for the development of later‐life cardiovascular disease in women. ¹ In fact, a meta‐analysis of 22 studies has reported that a history of preeclampsia is associated with a 4‐fold increased risk of heart failure and a 2‐fold increased risk of stroke. ² Importantly, these higher risks occur early in life, as a recent life‐trajectory study found that the risk of an adverse cardiovascular event in women who had preeclampsia, compared with women who had uncomplicated pregnancies, is at its highest 7 to 10 years postpartum, but continues to be elevated up to 20 years after delivery. ³ Two hypotheses have been proposed to explain the increased long‐term cardiovascular risk among women with a history of preeclampsia; (1) preeclampsia unveils preexisting cardiovascular disease conditions in women, including those with subclinical risk factors before conception, and (2) preeclampsia itself leads to persistent vascular dysfunction and aggravates preexisting subclinical risk factors for cardiovascular disease. ⁴ While there is evidence supporting both hypotheses, and they are not mutually exclusive, the underlying mechanisms for the increased cardiovascular risk in women who had preeclampsia when they were pregnant are still not fully understood.

Hypercholesterolemia is a major risk factor for the development of cardiovascular disease, ⁵ while also being a physiological process during pregnancy that is required for normal fetal development. ⁶ , ⁷ In contrast, excessive hypercholesterolemia in pregnancy predisposes the mother to the development of vascular complications in pregnancy, such as preeclampsia. ⁸ Similarly, preclinical studies (in mice and rats) have shown that excessive hypercholesterolemia in pregnancy, using a high‐cholesterol diet (HCD), creates a preeclampsia‐like phenotype that includes high blood pressure, proteinuria, reduced placental efficiency, and vascular dysfunction. ⁹ , ¹⁰ , ¹¹ , ¹² , ¹³ , ¹⁴ Outside of pregnancy, hypercholesterolemia impairs vascular function by reducing NO production via downregulation of endothelial NO synthase (eNOS), ¹⁵ while also increasing NO scavenging via superoxide, ¹⁶ , ¹⁷ leading to the formation of peroxynitrite. NO is a potent endothelium‐derived vasodilator ¹⁸ that helps to reduce peripheral vascular resistance and, consequently, lowers blood pressure. It is well accepted that women with preeclampsia have impaired NO signaling in pregnancy, and a meta‐analysis of 37 studies has suggested that this alteration in the NO pathway may persist later in life. ¹⁹ Interestingly, vascular endothelial dysfunction, secondary to changes in the NO pathway, increases arterial stiffness. ²⁰ This is relevant, as clinical studies using noninvasive ultrasound‐based techniques have shown that women who experienced preeclampsia have increased carotid artery stiffness ²¹ and thickness ²² , ²³ , ²⁴ , ²⁵ , ²⁶ (ie, a marker of subclinical atherosclerosis); however, the mechanisms causing this increased vascular stiffness have yet to be identified. Changes in the extracellular matrix, including alterations in elastin density and collagen deposition, can induce arterial stiffening in conditions of reduced NO availability (such as aging ²⁷ ), but whether such a process also underlies arterial stiffening in women with a history of preeclampsia, particularly in the context of excessive hypercholesterolemia in pregnancy, is not known.

We have previously shown that mice exposed to excessive hypercholesterolemia during pregnancy still have high circulating total cholesterol levels 3 months after pregnancy. ²⁸ In addition, these mice presented with impaired endothelial function in abdominal aortas postpartum, due to a reduction in the NO contribution to vasodilation together with an increase in superoxide levels and peroxynitrite formation. ²⁸ Of note, peroxynitrite can activate prostaglandin H synthase (PGHS), ²⁹ the enzyme that produces prostaglandin H₂, a vasoactive molecule that can be converted into prostacyclin (a vasodilator) or thromboxane A₂ (a vasoconstrictor). ³⁰ However, whether a history of excessive hypercholesterolemia during pregnancy, which is estimated to affect ≈25% of all pregnancies, ³¹ , ³² , ³³ , ³⁴ impacts vascular function in other vascular beds, as well as vascular compliance, is also not known. In the current study, we evaluated whether excessive hypercholesterolemia in pregnancy, leading to signs of the pregnancy complication preeclampsia, impacts maternal vascular function in mesenteric arteries (systemic resistance arteries that are important for blood pressure regulation), coronary arteries (representative cardiac vessel), carotid arteries (vascular bed that directly supplies the cerebral vasculature), and thoracic aortas (major conduit artery) 3 months after pregnancy (equivalent to ≈7.5 years postpartum in human years ³⁵ ), as well as identify the underlying mechanisms. Age‐matched never‐pregnant groups were included to assess whether the effects of the HCD were specific to pregnancy. We hypothesized that excessive hypercholesterolemia in pregnancy is associated with long‐term alterations in the NO and PGHS pathways, which impairs maternal vascular function and compliance later in life.

Methods

The raw data supporting the findings of this study are available from the corresponding author upon reasonable request.

Ethics Statement

All animal procedures performed in this study were reviewed and approved by the Animal Care and Use Committee of the University of Alberta (AUP‐3692) and are in accordance with the Canadian Council on Animal Care Guidelines and Animal Research: Reporting of In Vivo Experiments reporting guidelines. ³⁶

Animal Model and Experimental Design

Sprague Dawley rats (12‐week‐old; total: 66 females and 12 males [for breeding purposes]) were purchased from Charles River Laboratories and were housed in the animal facility at the University of Alberta under standard light–dark cycles (14–10 hours) with ad libitum access to water and food (standard chow diet). Same‐sex animals were housed in pairs in individually ventilated cages containing a plastic tube, a wood chewing block, and 2 cotton pads. After 1 week (acclimation period), female and male rats were allowed to mate overnight, and the presence of sperm or a plug in a vaginal smear the next morning indicated that the female rat was pregnant (gestational day [GD] 0). Once pregnant, the female rat was weighted and single housed until the end of gestation. On GD6, dams were arbitrarily allocated to receive a control diet (CD; n=20) or HCD (n=22; Modified LabDiet 5001 with 2% cholesterol and 0.5% cholic acid; LabDiet, St. Louis, MO) until GD20, which results in a preeclampsia phenotype as previously described by our group and others. ⁹ , ¹⁰ , ¹¹ , ¹² , ¹³ Both the CD and HCD groups received a CD from GD21 until the end point (3 months after birth). Dams were allowed to give birth (term pregnancy=22 days), after which the offspring was removed to avoid lactation as a covariant. Thereafter, animals were housed in pairs. Never‐pregnant groups were included to assess whether the effects of the HCD were specific to pregnancy. For these groups, age‐matched virgin female Sprague Dawley rats received the same diet intervention (ie, CD or HCD; n=12 per group) for 14 days, and were then kept (on the CD) in the animal facility until 3 months after the diet. To minimize the impact of each batch of animals as a potential confounder, every experimental group was proportionally represented in every batch of animals used in this study. The week before euthanasia, blood pressure was assessed (see below). Three months postpartum (equivalent to ≈7.5 years postpartum in human years ³⁵ ), rats were euthanized by cardiac puncture under isoflurane anesthesia (4% in 100% O₂). The humane end points set for this study included weight loss >10% during pregnancy or if the animal was visibly ill, both of which were checked daily by the animal facility staff and weekly by the research team. Blood was collected in BD Vacutainer tubes coated with EDTA (BD Biosciences, Mississauga, ON, Canada), centrifuged at 2000g for 10 minutes at 4 °C, and the plasma fraction was separated and snap‐frozen for ELISAs. Mesenteric arteries, coronary arteries, carotid arteries (common, external, and internal), and aortas were isolated, placed in ice‐cold HEPES‐buffered physiological salt solution (PSS, in mmol/L: 142 NaCl, 4.7 KCl, 1.17 MgSO₄, 1.56 CaCl₂, 1.18 K₂PO₄, 10 HEPES, 5.5 glucose, pH 7.4), and used to assess ex vivo vascular function, snap‐frozen embedded in Tissue‐Tek OCT Compound (Sakura Finetek Inc., Torrance, CA) for cryosectioning, or snap‐frozen for western blotting. Of note, ex vivo vascular function was assessed in all vascular beds using wire myography, with the exception of the external carotid artery, which was assessed using pressure myography (see below). A partial blinding approach was used in this study. All animals were sequentially numbered upon arrival in the animal facility. During pregnancy or the diet intervention, one investigator (A.A.d.O.) knew which experimental group the rats were in, as they were applying the intervention to pregnant or nonpregnant rats. However, at the time of euthanasia/conducting the experiments at 3 months postpartum, and while analyzing the data, all investigators involved were kept unaware of the experimental group to which the animals belonged until after the experiments/analysis.

Blood Pressure Measurements

CD and HCD rats were trained daily for 1 week (30‐minute training sessions) in nose cone restrainers before blood pressure measurements using noninvasive tail‐cuff CODA High Throughput System (Kent Scientific, Torrington, CT). The reported systolic (SBP), diastolic, and mean arterial blood pressures are averages of at least 10 independent blood pressure cycles for each rat.

Cholesterol and Oxidized LDL Assays

Cholesterol (total, high‐density lipoprotein, and low‐density lipoprotein [LDL]/very low‐density lipoprotein) and oxidized LDL (oxLDL) levels were evaluated in CD and HCD plasma samples using the EnzyChrom high‐density lipoprotein and LDL/very low‐density lipoprotein assay kit (E2HL‐100; BioAssay Systems, Hayward, CA) and Rat OxLDL ELISA kit (MBS2501477; MyBiosource Inc., San Diego, CA), respectively. Samples were processed and analyzed according to the manufacturer's instructions and were used undiluted.

Ex Vivo Assessment of Vascular Function Using Wire Myography

Vascular function was assessed in second‐order mesenteric arteries, left anterior descending coronary arteries, left common and internal carotid arteries, and thoracic aortas using a wire myograph system (620M; Danish Myo Technology, Copenhagen, Denmark), as follows.

• Mesenteric, coronary, and common/internal carotid arteries (≈2 mm in length) were mounted on a wire myograph system, and a stepwise procedure was used to determine the optimal resting tension of each artery (0.8×IC100, 13.3 kPa). Then, mesenteric and carotid arteries were stimulated with phenylephrine (10 μmol/L; Sigma‐Aldrich, Oakville, ON, Canada), whereas coronary arteries were stimulated with 9,11‐Dideoxy‐9a,11a‐methanoepoxy prostaglandin F_2a (U46619, a thromboxane A₂ analog, 1 μmol/L; Sigma‐Aldrich) to induce vasoconstriction for 5 minutes. Vessels were then thrice washed and allowed to recover for 10 minutes. Next, vessels were restimulated with phenylephrine (10 μmol/L) or U46619 (1 μmol/L; only in coronary arteries) for 5 minutes, after which methacholine was added (3 μmol/L; Sigma‐Aldrich) for 2 minutes to assess endothelial integrity.

• Aortas (≈2 mm in length) were mounted on the wire myograph system and normalized by being stretched to a resting tension of 30 mN. During normalization (≈1 hour), the PSS buffer was replaced every 15 minutes and resting tension readjusted to 30 mN each time. Next, vessels were exposed to a high K⁺ PSS solution (123 mmol/L) for 15 minutes, thrice washed, and allowed to recover for 10 minutes. After that, vessels were stimulated with phenylephrine (1 μmol/L) for 15 minutes, and methacholine (3 μmol/L) was added for 2 minutes.

Thereafter, all arterial segments were thrice washed with PSS and allowed to return to baseline. Cumulative concentration response curves (CCRCs) to phenylephrine (1 nmol/L to 100 μmol/L) were obtained in mesenteric arteries, carotid arteries, and aortas, whereas CCRCs to U46619 (1 nmol/L to 10 μmol/L) were obtained in the coronary arteries. In addition, CCRCs to phenylephrine were obtained in carotid arteries incubated for 30 minutes with N(G)‐nitro‐l‐arginine methyl ester hydrochloride (L‐NAME, a pan‐NOS inhibitor, 100 μmol/L; Sigma‐Aldrich). Next, vessels were washed with PSS, allowed to recover to baseline, and CCRCs to methacholine (100 pmol/L to 100 μmol/L) or sodium nitroprusside (100 pmol/L to 10 μmol/L; Sigma‐Aldrich) were obtained in vessels preconstricted with phenylephrine or U46619 (coronary arteries). The preconstriction doses of phenylephrine for mesenteric arteries (5 μmol/L), carotid arteries (1 μmol/L), and aortas (1 μmol/L) were fixed for all experiments, as no differences were observed in the 80% maximal response of the drug in pilot studies, whereas the preconstriction dose (80% maximal response) of U46619 for coronary arteries was calculated for each vessel after the U46619 CCRC. CCRCs to methacholine were also obtained in common carotid arteries incubated for 30 minutes with L‐NAME (100 μmol/L), meclofenamate (a pan‐PGHS inhibitor, 1 μmol/L; Sigma‐Aldrich), or NS398 (a selective prostaglandin H synthase isoform 2 [PGHS2] inhibitor, 1 μmol/L; Tocris Biosciences, Burlington, ON, Canada). This was the end of the protocol for the aortas, while the mesenteric, coronary, and carotid arteries were washed with PSS, allowed to recover for at least 15 minutes, and challenged with a high K⁺ PSS solution (123 mmol/L) for 10 minutes to finalize the experiments.

All vascular data were normalized to vessel length, and are summarized as maximum responses (E_max), the response at the highest concentration of the agonist, the sensitivity to the agonist (pEC₅₀; the negative log of the drug concentration that produces 50% of the maximum response), or as the difference between curves with and without inhibitors (delta E_max; vehicle versus L‐NAME, to determine the NO contribution to vasodilation or NO modulation of vasoconstriction). Vasoconstriction data were calculated as force displacement from baseline (mN/mm), while vasodilation data were calculated as percentage of vasodilation relative to the maximum constriction induced by phenylephrine (in mesenteric and carotid arteries, as well as aortas) or U46619 (in coronary arteries).

Ex Vivo Assessment of Vascular Function Using Pressure Myography

The functional and mechanical properties of the external carotid arteries (a collateral blood vessel for the cerebral circulation ³⁷ , ³⁸ ) were assessed using a pressure myograph system (114P, Danish Myo Technology), as previously described by our group. ¹³ , ³⁹ In brief, arterial segments (≈3–4 mm) were cannulated on each side with 2 glass micropipettes, and the intraluminal pressure of the vessel was increased in a stepwise manner (10 mm Hg increments with 5‐minute intervals) from 4 to 60 mm Hg, at which pressure the arteries were left for 30 minutes. Then, the intraluminal pressure was decreased to 4 mm Hg, and the active properties of the vessel were assessed by increasing the intraluminal pressure of the vessel from 4 to 160 mm Hg (10 mm Hg increments with 2‐minute intervals). Thereafter, the intraluminal pressure was returned to baseline, arteries were thrice washed and incubated with a Ca²⁺ free PSS containing EGTA (2 mmol/L; Sigma‐Aldrich) and papaverine (0.1 μmol/L; Sigma Aldrich). After 20 minutes, the passive properties of the vessel were assessed by again increasing the intraluminal pressure of the vessel from 4 to 160 mm Hg (10 mm Hg increments with 2‐minute intervals). Throughout the experiment, the intraluminal pressure and vessel internal–outer diameters were measured using an inverted microscope (Axio Vert A.1, Zeiss, Germany) coupled to the MyoVIEW4 software (Danish Myo Technology). Wall thickness, lumen diameter, wall thickness‐to‐lumen diameter ratio, and cross‐sectional area of the arteries were determined at baseline (4 mm Hg) before the assessment of the passive properties of the vessel. In addition, pressure‐dependent changes in the internal diameter and wall thickness of the vessel during the active and passive curves were used to calculate the percentage of myogenic tone development, as well as circumferential stress and strain. ³⁹ , ⁴⁰ Since the circumferential stress and strain relationship is nonlinear, the incremental elastic modulus was calculated by fitting the resulting circumferential stress and strain curve to an exponential equation to derive the β coefficient, a marker of arterial stiffness, as described elsewhere. ⁴¹ , ⁴²

Western Blotting for Protein Expression of NOS and PGHS

The Pierce BCA Protein Assay kit (Thermo Fisher Scientific, Waltham, MA) was used to determine the total protein concentration in homogenized common carotid artery samples from CD and HCD postpartum rats. A total of 50 μg of protein were separated on either 6% or 8% SDS‐polyacrylamide gels (based on the predicted molecular weight of the protein of interest) and were subsequently transferred to 0.2‐μm nitrocellulose membranes (Bio‐Rad, Hercules, CA). Next, membranes were stained with the Li‐Cor Revert 700 Total Protein Stain (Li‐Cor Biosciences, Lincoln, NE) following the methods on the manufacturer's pack insert, and visualized with the Li‐Cor Odyssey Imaging System (version 3.0; Li‐Cor Biosciences). Afterwards, membranes were blocked with Blocking Buffer for Fluorescent Western Blotting (Rockland Immunochemicals, Pottstown, PA) for 1.5 hours, and probed overnight at 4 °C with primary antibodies for eNOS (1:500; mouse monoclonal; BD Biosciences), inducible NOS (1:500; mouse monoclonal; BD Biosciences), PGHS isoform 1 (1:5000; rabbit polyclonal; Abcam, Waltham, MA), or PGHS2 (1:250; rabbit polyclonal; Abcam) diluted in 1× PBS with 0.1% Tween 20. The next day, membranes were thrice washed in 1× PBS with 0.1% Tween 20 for 10 minutes, incubated with their corresponding secondary antibodies: IRDye 800CW donkey anti‐mouse IgG (eNOS and inducible NO synthase) and IRDye 800CW donkey anti‐rabbit IgG (PGHS isoform 1, and PGHS2) for 1 hour at 1:10 000 dilution in 25% Blocking Buffer and 1× PBS with 0.1% Tween 20, then thrice washed in 1× PBS with 0.1% Tween 20 for 10 minutes, and visualized with the Li‐Cor Odyssey Imaging System (Li‐Cor Biosciences). ImageStudioLite software (Li‐Cor Biosciences) was used to determine band densitometry, and the reported values for each protein of interest were normalized to the total protein staining (using Li‐Cor Revert 700), and are presented as a percentage of the averaged CD group.

Detection of Nitrotyrosine and ROS

Common carotid artery sections (thickness, 8 μm) from CD and HCD postpartum rats were assessed using immunofluorescence to detect nitrotyrosine (1:50, mouse antinitrotyrosine; Novus Biologicals, Oakville, ON, Canada) or dihydroethidium fluorescent staining to detect ROS levels (Biotium Inc., Fremont, CA), as previously described by our group. ¹³ Images were acquired with an Olympus IX81 Microscope (Olympus Canada Inc., Toronto, ON, Canada), and analyzed with FIJI software version 2.14.0/1.54f. ⁴³ The reported mean fluorescent intensity (arbitrary units) was calculated by highlighting the total vessel area, excluding the adventitia and nonspecific staining or folded areas. Representative images were adjusted for brightness and contrast using the same parameters.

Histological Assessment of Arterial Remodeling

Sections (thickness, 8 μm) of common carotid arteries of CD and HCD postpartum rats were formalin‐fixed for 10 minutes and stained with hematoxylin (HT‐107; Sigma‐Aldrich) and counterstained with eosin (HT‐110116; Sigma‐Aldrich) to assess the carotid intima media thickness. Collagen was assessed with a Trichrome Stain (Masson) kit (HT‐15; Sigma‐Aldrich) following the manufacturer's guidelines, while elastin was assessed with an Elastic Stain kit (HT‐25A; Sigma‐Aldrich) also following the manufacturer's instructions but without the addition of the Van Gieson solution for counterstaining of collagen fibers. Slides were imaged using a bright field microscope (EVOS XL Core Imaging System; ThermoFisher Scientific), and analyzed with the FIJI software. ⁴³ In brief, line tracing was used to determine the carotid intima media thickness, and the reported values are averages of 3 independent measurements for each image. For the collagen and elastin staining, color deconvolution tools were used to assess collagen deposition or elastin density in the tunicae media and intima, and the reported data are presented as the percentage of positive area. In all cases, nonspecific staining or folded areas were excluded from the analyses. Representative images were adjusted for brightness and contrast using the same parameters.

Experimental Exclusions

All CD and HCD postpartum/never‐pregnant rats designated for this project were included in our analysis, but not all measured outcomes were included in the reported data. One oxLDL data point in the CD postpartum group was excluded as a statistical outlier (see below). In the common carotid arteries of the postpartum rats, 2 CCRCs to phenylephrine in the presence of L‐NAME for the CD group were excluded as there was constriction following the addition of L‐NAME to the bath. One CCRC to methacholine in the presence of NS398 for the CD and another in the HCD group were removed as they were statistical outliers (see below). One myogenic tone curve and 1 circumferential stress and strain curve for the HCD group were removed due to the presence of arterial side branches, while another circumferential stress and strain curve for the HCD group was removed as a statistical outlier (see below). One DHE data point in the HCD group and 1 nitrotyrosine data point in the CD and another in the HCD postpartum groups were excluded as the sections detached from the slide.

Statistical Analysis

Sample size was estimated based on our previous results using a similar study design to assess vascular function (our primary end point), ¹³ , ²⁸ under the assumption of 80% power with an α error of 0.05, considering estimates of effect size measures (ie, Cohen's d) and a standard deviation of 20% of the mean. Prism software version 10.0.3 (GraphPad Software, San Diego, CA) was used to statistically analyze data, which are shown as means±SEM. The Grubb's test was used to identify outliers, which were removed from the statistical analyses. The D'Agostino and Pearson test was used to determine if the data are normally distributed. Data were analyzed with an unpaired t test (if both populations had the same SD) followed by an F test or an unpaired t test with Welch's correction (if no equal SD) or a 2‐way ANOVA with a subsequent Sidak's post hoc test; P≤0.05 was considered statistically significant, and n represents the number of rats used per group.

Results

Exposure to an HCD in Pregnancy Did Not Affect the Maternal Lipid Profile and Blood Pressures Later in Life, But Increased the Variance of Plasma oxLDL Concentrations and SBP

No changes in litter size were observed between the CD and HCD groups on postnatal day 1 (13.35±0.48 versus 12.95±0.57, n=20–22). The maternal body weight, lipid profile (total cholesterol, high‐density lipoprotein, LDL, and oxLDL), and blood pressures (SBP, diastolic blood pressure, and mean arterial blood pressure) were also not different between the CD and HCD postpartum groups 3 months after the pregnancy (Table). Similarly, no differences were observed in these parameters between the CD and HCD groups of age‐matched never pregnant female rats (Table S1). Interestingly, rats within the HCD group had more variation in their plasma oxLDL concentrations (range CD, 0.211 versus HCD, 2.519 pg/μL; P=0.004) and SBP (range CD, 15.33 versus HCD, 61.93 mm Hg; P=0.0324) compared with rats within the CD group (Table). This variation was not observed in the age‐matched never‐pregnant female rat groups (Table S1).

Table 1.

Body Weight, Plasma Lipid Profile, and Blood Pressures in CD and HCD Rats 3 Months Postpartum

Outcome	CD	HCD	Unpaired t test
Body weight, g	383.1±10.1 (n=20)	383.3±10.0 (n=22)	ns
Total cholesterol, mg/dL	74.90±9.44 (n=9)	88.82±7.93 (n=9)	ns
LDL/VLDL, mg/dL	11.38±2.12 (n=9)	11.17±1.63 (n=9)	ns
High‐density lipoprotein, mg/dL	69.96±10.35 (n=9)	76.20±6.51 (n=9)	ns
Oxidized LDL, pg/μL	2.42±0.04 (n=4)	2.00±0.40 (n=5)	ns
Systolic blood pressure, mm Hg	121.50±2.61 (n=5)	127.90±6.22 (n=10)	ns
Diastolic blood pressure, mm Hg	88.24±3.12 (n=5)	89.08±3.76 (n=10)	ns.
Mean arterial blood pressure, mm Hg	99.03±2.84 (n=5)	101.70±4.43 (n=10)	ns.

Data are presented as means±SEM, and were analyzed using an unpaired t test followed by an F test or an unpaired t test with Welch's correction. CD indicates control diet; HCD, high‐cholesterol diet; LDL low‐density lipoprotein; ns, not significant; and VLDL, very‐low‐density lipoprotein.

Vascular Function Was Impaired in Aortas and Carotid Arteries but Not in Mesenteric and Coronary Arteries of HCD Postpartum Rats

Vasoconstriction responses to phenylephrine (mesenteric arteries) or U46619 (coronary arteries) were similar between the CD and HCD rats 3 months after pregnancy (Figure 1A and 1B). In contrast, maximum vasoconstriction responses to phenylephrine were reduced in aortas and increased in common carotid arteries from HCD postpartum rats compared with CD controls (Figure 1C and 1E). In the internal carotid arteries, maximum vasoconstriction responses to phenylephrine were similar between the CD and HCD postpartum groups, whereas the sensitivity (pEC₅₀) to phenylephrine was higher in the HCD postpartum group (Figure 1D). No effects of the diet were observed in vasoconstriction responses to phenylephrine in the age‐matched never‐pregnant female rats (Figure 1F and Figure S1). Vasoconstriction responses to high K⁺ PSS were similar between the CD and HCD postpartum groups in all vascular beds assessed (Figure S2).

Figure 1. HCD in pregnancy impaired vascular responses to phenylephrine in aortas and carotid arteries 3 months postpartum.

Vasoconstriction to phenylephrine (i) and U46619 (i; only in B) and E_max (ii) or pEC₅₀ (only in D, ii) in mesenteric arteries (A), coronary arteries (B), thoracic aortas (C), internal carotid arteries (D), and common carotid arteries (E) of CD (solid squares) and HCD (open squares) rats 3 months after the pregnancy. F, Vasoconstriction to phenylephrine (i) and E_max (ii) in common carotid arteries of age‐matched never‐pregnant CD (blue solid triangles) and HCD (blue open triangles) female rats 3 months after the diet intervention. Data are shown as means±SEM; *P<0.05 using an unpaired t test, n=4–10 rats/group. CD indicates control diet; E_max, maximum response; HCD, high‐cholesterol diet; PE, phenylephrine; and pEC₅₀, negative log of the drug concentration that produces 50% of the maximum response.

Endothelium‐dependent vasodilation responses to methacholine were similar between the CD and HCD postpartum groups in the mesenteric arteries, coronary arteries, aortas, and internal carotid arteries (Figure 2A through 2D). However, maximum vasodilation to methacholine was reduced in common carotid arteries of the HCD postpartum rats compared with the CD controls (Figure 2E). No effects of the diet were observed in vasodilation to methacholine in the age‐matched never‐pregnant female rats (Figure 2F). Moreover, endothelium‐independent vasodilation to sodium nitroprusside was similar between the CD and HCD postpartum groups (Figure S3).

Figure 2. HCD in pregnancy reduced maximum endothelium‐dependent vasodilation to methacholine only in carotid arteries 3 months postpartum.

Vasodilation to methacholine (i) and E_max (ii) in mesenteric arteries (A), coronary arteries (B), thoracic aortas (C), internal carotid arteries (D), and common carotid arteries (E) of CD (black solid squares) and HCD (black open squares) rats 3 months after the pregnancy. (F) Vasodilation to methacholine (i) and E_max (ii) in common carotid arteries of age‐matched never pregnant CD (blue solid triangles) and HCD (blue open triangles) female rats 3 months after the diet intervention. Data are shown as means±SEM; *P<0.05 using an unpaired t test, n=4–12 rats/group. CD indicates control diet; E_max, maximum response; HCD, high‐cholesterol diet; and MCh, methacholine.

HCD in Pregnancy Altered the NO and PGHS Pathways in Carotid Arteries Later in Life

As the HCD primarily affected the carotid arteries (vasoconstriction to phenylephrine in common and internal carotid arteries and vasodilation to methacholine in common carotid arteries) of postpartum rats, the following mechanistic studies were conducted exclusively in this vascular bed. L‐NAME (a pan‐NOS inhibitor) significantly increased vasoconstriction responses to phenylephrine in common carotid arteries of the CD and HCD postpartum rats, but this NO modulation of phenylephrine‐induced constriction was less in the HCD group compared with CD (delta E_max; Figure 3A). Similarly, in the internal carotid arteries, pan‐NOS inhibition increased the sensitivity (pEC₅₀) to phenylephrine in both postpartum groups, abolishing the differences in the pEC₅₀ of the drug between the CD and HCD postpartum groups (direct comparison of the L‐NAME curves; Figure S4), which suggests that the NO modulation of phenylephrine‐induced constriction is reduced in the HCD postpartum group compared with CD. In addition, preincubation of the common carotid arteries with L‐NAME prevented vasodilation to methacholine in the CD and HCD postpartum rats, but this effect was more pronounced in the HCD group compared with CD rats (delta E_max; Figure 3B). No differences were observed in the protein expression levels of eNOS and inducible NO synthase, as well as in the levels of ROS and nitrotyrosine between the groups in common carotid arteries (Figure 3C through 3E).

Figure 3. HCD in pregnancy altered the NO pathway in carotid arteries 3 months postpartum.

A, Vasoconstriction to phenylephrine (i) and delta E_max (vehicle vs inhibitor, ii) and (B) vasodilation to methacholine (i) and delta E_max (vehicle vs inhibitor, ii) in common carotid arteries of CD (solid squares) and HCD (open squares) rats 3 months after the pregnancy. Vessels were incubated with vehicle (black) or L‐NAME (a pan‐NOS inhibitor, red). C, eNOS blot and densitometry and (D) iNOS blot and densitometry. E, Representative images (i) and densitometry analysis of ROS (ii) and nitrotyrosine (iii). Scale bar=100 μm. Data are shown as means±SEM; *P<0.05 using an unpaired t test, n=4–10 rats/group. a.u. indicates arbitrary units; CD, control diet; E_max, maximum response; eNOS, endothelial NO synthase; HCD, high‐cholesterol diet; iNOS, inducible NO synthase; L, protein ladder; L‐NAME, N(G)‐nitro‐l‐arginine methyl ester hydrochloride; MCh, methacholine; MFI, mean fluorescence intensity; NOS, NO synthase; and ROS, reactive oxygen species.

As stimulation of endothelial cells with methacholine can also activate the PGHS pathway, vasodilation experiments were performed with and without meclofenamate (a pan‐PGHS inhibitor) and NS398 (a selective PGHS2 inhibitor) in the common carotid arteries. Pan‐inhibition of PGHS with meclofenamate improved vasodilation to methacholine in the common carotid arteries of the HCD postpartum group, without impact in the carotid arteries of the CD rats (Figure 4A). Likewise, NS398 improved vasodilation to methacholine in the HCD postpartum rats, while no effects were observed on the carotid arteries of the CD rats (Figure 4B). However, no differences were observed in the protein expression levels of PGHS isoform 1 and PGHS2 between the CD and HCD postpartum groups (Figure 5C).

Figure 4. Inhibition of PGHS2 improved endothelium‐dependent vasodilation to methacholine in carotid arteries of the HCD group 3 months postpartum.

Vasodilation to methacholine (i) and vasodilation at the highest methacholine concentration (E_max; ii) in common carotid arteries of CD (solid squares) and HCD (open squares) postpartum rats incubated with vehicle (black) or meclofenamate (pan‐PGHS inhibitor, orange; A) and vehicle (black) or NS398 (selective PGHS2 inhibitor, magenta; B). C, PGHS1 blot and densitometry (i and iii) and PGHS2 blot and densitometry (ii and iv) in common carotid arteries. Data are shown as means±SEM; *P<0.05, **P<0.01, ***P<0.001 using 2‐way ANOVA with Sidak's post hoc test, n=5–10 rats/group. CD indicates control diet; E_max, maximum response; HCD, high‐cholesterol diet; L, protein ladder; MCh, methacholine; PGHS, prostaglandin H synthase; PGHS1, prostaglandin H synthase 1; and PGHS2, prostaglandin H synthase 2.

Figure 5. HCD in pregnancy increased carotid artery stiffness 3 months postpartum.

A, Wall thickness (i), lumen diameter (ii), wall thickness‐to‐lumen diameter ratio (iii), and CSA (iv) in external carotid arteries of CD (solid squares) and HCD (open squares) postpartum rats. B, Circumferential stress–strain curve (i) and AUC for stress (ii) and strain (iii). (C) Myogenic tone (i) and AUC (ii). Representative images (i) and densitometry analysis (ii) of Trichrome stain (Masson) (D; collagen fibers are stained in blue, black arrows) and Verhoeff stain (E; elastin fibers are stained dark, black arrows) in common carotid arteries of CD and HCD postpartum rats. Scale bar=100 μm. Image analysis of collagen and elastin was performed specifically in the tunicae intima and media (see Methods for details). Data are shown as means±SEM; *P<0.05 using an unpaired t test, n=5–10 rats/group. AUC indicates area under the curve; CD, control diet; CSA, cross‐sectional area; and HCD, high‐cholesterol diet.

Carotid Artery Stiffness Was Increased in HCD Rats Postpartum and Associated With Reduced Elastin Density

The lumen diameter, wall thickness, wall thickness‐to‐lumen diameter ratio, and cross‐sectional area (at 4 mm Hg) of the external carotid arteries were similar 3 months postpartum between the rats exposed to CD or HCD in pregnancy (Figure 5A). However, the circumferential strain but not circumferential stress was less in the HCD postpartum group compared with CD rats (Figure 5B). In addition, the β coefficient of the incremental elastic modulus, a marker of arterial stiffness, was increased in the HCD postpartum rats compared with CD (3.84±0.3 versus 2.84±0.12; P=0.0038). In addition, in response to increasing pressure, carotid arteries of HCD postpartum rats developed less myogenic tone compared with the CD controls (Figure 5C). No effects of the HCD on artery stiffness and myogenic tone development were observed in the age‐matched never pregnant female rats (Figure S5).

As the main determinants of vascular wall structure, collagen and elastin levels were evaluated in common carotid arteries of the CD and HCD rats postpartum. Collagen deposition was similar between the CD and HCD groups (Figure 5D), whereas elastin density was lower in the HCD postpartum group compared with CD (Figure 5E). In addition, no differences were observed in the carotid intima media thickness (Figure S6), marker of subclinical atherosclerosis, between the CD and HCD groups 3 months postpartum.

Discussion

In this study, we demonstrated that rats on an HCD, only in pregnancy, presented with impaired vascular function in carotid arteries and aortas, but not in mesenteric and coronary arteries, 3 months after the pregnancy (equivalent to ≈7.5 years postpartum in human years ³⁵ ). In addition, exposure to an HCD in pregnancy increased carotid artery stiffness and reduced the intrinsic ability of this blood vessel to develop pressure‐dependent myogenic tone later in life. Importantly, all of the vascular changes observed in the HCD postpartum rats were absent in the age‐matched never‐pregnant female rats that received the same HCD intervention, indicating that there are long‐term pregnancy‐specific alterations in the maternal vascular system following a pregnancy complicated by excessive hypercholesterolemia. These adaptations in the maternal carotid arteries following a pregnancy complicated by excessive hypercholesterolemia (a significant risk factor for the development of preeclampsia ⁸ ) may explain the increased risk of cardiovascular events (such as stroke) in women with a history of preeclampsia. ²

Excessive hypercholesterolemia in pregnancy did not affect the body weight or lipid profile 3 months after pregnancy in rats. This is partially in line with our previous study in which mice that were fed a hypercholesterolemic diet during late pregnancy also did not present with changes in body weight later in life but had sustained increased total cholesterol levels ²⁸ and may reflect species‐specific mechanisms in lipid metabolism and clearance. Of note, conflicting results regarding the maternal lipid profile have also been reported in humans. While higher total cholesterol and LDL cholesterol levels have been reported in women who had a pregnancy complicated by preeclampsia compared with women who had uncomplicated pregnancies in a 1‐year follow‐up study, ⁴⁴ another study did not identify differences in the lipid profile of women who had experienced preeclampsia compared with women who had experienced normotensive pregnancies 6 months and 2 years earlier. ⁴⁵ This may be explained by the fact that both studies lacked specific data regarding the maternal lipid profile before pregnancy. Thus, it is possible that the observed postpartum levels in the preeclampsia group reflect the pre‐pregnancy status of these women. In the present study, we also showed that exposure to excessive hypercholesterolemia in rat pregnancy did not affect maternal blood pressures 3 months after the pregnancy. This was likely due to preserved vascular function in the mesenteric arteries (see below). Conversely, a previous study has demonstrated an increased SBP, but not diastolic blood pressure, in postpartum rats (5 months postpartum) that experienced a pregnancy complicated by excessive hypercholesterolemia. ⁴⁶ This is, however, a model in which the rats were maintained on an HCD throughout the postpartum period, and thus differs from our insult, which was only given during pregnancy (from GD6 to GD20). Interestingly, we observed an increase in oxLDL and SBP variability within the HCD rats compared with the CD rats, meaning that the impact of a hypercholesterolemic diet in pregnancy on maternal oxLDL levels and SBP later in life varies among HCD rats. This could be explained by a similar variation in the severity of the preeclampsia‐like phenotype developed at the end of gestation. ¹³ Together, these data suggest that having had a pregnancy complicated by excessive hypercholesterolemia in itself has long‐term implications for the maternal vascular system, which are likely not due to persistent changes in the maternal lipid profile at 3 months postpartum or increased blood pressure (independent risk factors for the development of cardiovascular disease). ⁵

Exposure to excessive hypercholesterolemia in pregnancy reduced vasodilation to methacholine in common carotid arteries but not in mesenteric arteries, coronary arteries, or aortas 3 months postpartum. This is interesting because while all vascular beds were exposed to the same insult in pregnancy, there was a clear vascular bed–specific impact later in life. Our present findings in thoracic aortas of HCD postpartum rats differ from our previous study in abdominal aortas of HCD postpartum mice. ²⁸ This disparity may be explained by the fact that total cholesterol levels remain elevated in HCD postpartum mice but not in HCD postpartum rats, as well as by the regional heterogeneity (ie, thoracic versus abdominal aortas) of endothelium‐dependent vasodilation in this vascular bed or by species‐specific differences. ⁴⁷ , ⁴⁸ Differences in vascular responses between arterial beds in postpartum rats have also been reported in the reduced uteroplacental perfusion rat model of preeclampsia in which the mesenteric arteries had impaired vasodilation later in life, whereas no differences were observed between the groups in isolated aortas. ⁴⁹ Still, the issue of increased susceptibility of 1 vascular bed in relation to another persists, and calls for additional studies to dissect the impact of pregnancy‐specific insults on later‐life maternal vascular function in multiple vascular beds, which is of particular interest because preeclampsia is a multisystem disorder. ⁵⁰

While the difference in vasodilation to methacholine between the carotid arteries of the CD and HCD postpartum groups seems small (≈15%), there is a marked impact of lumen diameter in blood flow resistance. ⁵¹ In fact, according to Poiseuille's Law, resistance ($R$) is inversely proportional to the radius ($r$) of the blood vessel raised to the fourth power ($R=\frac{1}{r^4}$), which means that even small reductions in lumen diameter greatly increase blood flow resistance. This is important as the bilateral carotid arteries are the main supplier of blood to the cerebral circulation (≈75%), ⁵² and impaired carotid artery function may affect the cerebral blood flow autoregulation. ²² An interesting aspect of the cerebral circulation is the fact that both the small and large arteries (including extracranial blood vessels) supplying the brain are sites of vascular resistance. The greater contribution of the large vessels to the control of blood flow in the brain, compared with the peripheral vessels, may be an adaptive mechanism to maintain a continuous supply of oxygen and nutrients to this critical organ (reviewed in Cipolla ⁵³ ). Thus, it may be suggested that the impaired carotid artery function observed in this study can help to explain the higher risk of cerebrovascular events in women with a history of preeclampsia. ²

Further mechanistic studies were conducted exclusively on the common carotid artery as the vascular bed with impaired endothelium‐dependent vasodilation. The response to sodium nitroprusside, an endothelium‐independent vasodilator, was similar between the CD and HCD postpartum groups, suggesting the impairment in vasodilation was endothelial in origin. Carotid arteries are conduit vessels that rely almost entirely on NO for vasodilation. Thus, a potential role for NO in the impaired vasodilation to methacholine in carotid arteries was investigated using the pan‐NOS inhibitor, L‐NAME. As expected, pan‐inhibition of NOS significantly prevented vasodilation in carotid arteries from both postpartum groups. However, in the presence of L‐NAME, there was a residual dilation in the CD postpartum group, which was not observed in the HCD rats. Thus, there was an increased NO dependency of vasodilation in the carotid arteries of these rats. However, our molecular data did not support an increase in NO bioavailability in the HCD group compared with CD (see below). Stimulation of endothelial cells with methacholine can also lead to the activation of PGHS, the enzyme that produces prostaglandin H₂, which is the substrate for the formation of prostacyclin (vasodilator) and thromboxane A₂ (vasoconstrictor). Therefore, we assessed the potential contribution of PGHS. We observed that both pan‐inhibition of PGHS and selective inhibition of PGHS2 improved vasodilation to methacholine in the HCD group, suggesting that the impaired vasodilation was PGHS2 mediated. Interestingly, and contrary to our expectations, the protein expression levels of both PGHS2 and PGHS isoform 1 were similar between the groups, indicating that an HCD in pregnancy had long‐term effects on PGHS2 activity but not on PGHS2 protein expression. Thus, it seems that the impaired vasodilation to methacholine in the carotid arteries of HCD postpartum rats later in life is not directly mediated via alterations in the NO pathway but rather via activation of PGHS2, leading to the release of endothelium‐derived contractile factors such as thromboxane A₂.

While no differences were observed in mesenteric and coronary arteries, there was a contrasting impact of the HCD exposure in pregnancy on vasoconstriction responses to phenylephrine in the carotid arteries (ie, increased E_max in common carotid arteries and increased pEC₅₀ in internal carotid arteries) versus aortas (ie, reduced E_max). In none of the vascular beds studied were changes in the basal vascular smooth muscle cell function observed, as reflected by the non–receptor‐mediated constriction to high K⁺ PSS. This suggests that the detected changes in carotid arteries and aortas are likely due to alterations in mechanisms that are associated with α₁‐adrenergic stimulation and may include endothelial modulation of vasoconstriction; thus, a potential role for NO was assessed using L‐NAME. The increased capacity of the common carotid arteries to constrict in response to phenylephrine, as well as the increased sensitivity to the drug in internal carotid arteries, may be attributed to a reduction in the NO modulation of phenylephrine‐induced constriction in the HCD rats compared with CD. It is well known that NO plays a critical role in controlling basal vascular smooth muscle tone and that reductions in NO availability or signaling can result in increased vasoconstriction responses. However, contrary to our expectations, eNOS and inducible NO synthase protein expression levels, as well as ROS and nitrotyrosine levels (a footprint marker of peroxynitrite formation, the product of the reaction between superoxide and NO), were similar between the CD and HCD postpartum rats. Still, NOS activity (which determines NO availability) can be affected by other mechanisms, such as an increase in arginase 2 (an enzyme that competes with NOS for the substrate l‐arginine), ⁵⁴ a reduction in cationic aminoacid transporters (which are responsible for l‐arginine uptake), ⁵⁵ and posttranslational modifications of eNOS. ⁵⁶ With respect to the blunted responses to phenylephrine in thoracic aortas from the HCD postpartum rats, this is in line with our previous study, in which abdominal aortas of HCD postpartum mice also displayed a tendency toward a reduced capacity to constrict to phenylephrine later in life. ²⁸ A potential mechanism for the impaired response of the aortas from HCD postpartum rats is early vascular aging. Aging is associated with a decline in vascular responses to phenylephrine in male aortas, ⁵⁷ , ⁵⁸ and signs of early vascular aging have been reported in aortas from women who had preeclampsia. ⁵⁹ Indeed, the maximum vasoconstriction responses to phenylephrine in the CD postpartum group were not different than that of the CD and HCD age‐matched never pregnant female rats (ie, the reduced capacity to constrict was only observed in the HCD postpartum rats). Still, additional studies are required to directly test this hypothesis. Overall, our vasoconstriction data suggest that exposure to excessive hypercholesterolemia in pregnancy impairs later‐life maternal vasoconstriction responses to phenylephrine in carotid arteries and aortas and that the underlying mechanisms for these impairments are specific to each vascular bed.

Our data showed that exposure to excessive hypercholesterolemia in pregnancy primarily impacts the maternal carotid arteries later in life. Therefore, we also used pressure myography to further characterize the vascular function of these vessels. We found that exposure to excessively high cholesterol levels in pregnancy was associated with stiffer carotid arteries 3 months postpartum. In fact, the circumferential strain (ie, deformation of the vessel wall due to stress) but not circumferential stress (ie, force applied to the vessel wall per unit area) of the carotid arteries was reduced (indicating reduced compliance), while the β coefficient of the incremental elastic modulus (a marker of arterial stiffness) was increased, in HCD postpartum rats compared with CD. Carotid artery stiffness is increased in women with a history of preeclampsia compared with women who had a history of only uncomplicated pregnancies (6 months to 5 years postpartum), ²¹ but studies addressing the underlying molecular mechanisms lack in the literature. Diminished elastin density has been reported in aortas of HCD rats at the end of gestation, ⁶⁰ and changes in vascular wall extracellular matrix composition are at the core of increased arterial stiffness. Indeed, in our study, the increased carotid artery stiffness in the HCD postpartum rats could be attributed to a decrease in elastin density. In addition, no differences were observed between the postpartum CD and HCD groups in collagen deposition in the tunica media or in vessel wall remodeling (as a function of lumen diameter, wall thickness, wall thickness‐to‐lumen diameter ratio, and cross‐sectional area). We have previously observed that maternal blood vessels that have less elastic fibers than their counterparts are more stiff, despite no changes in collagen deposition. ³⁹ , ⁴⁰ This is because elastic fibers, particularly in large conduit arteries such as the carotid arteries, play a critical role in facilitating vascular compliance, as they provide energy storage and the ability to recoil in each cardiac cycle, ⁶¹ which dampens pulsatile blood flow for distal arteries (such as the resistance‐size cerebral arteries). Importantly, we also observed that the development of myogenic tone was lower in the HCD postpartum rats compared with CD rats. Various mechanisms may affect myogenic tone development, including reduced vessel compliance (ie, increased stiffness). ⁶² The myogenic response functions as an autoregulatory process to maintain constant blood flow to vital organs, ⁶² including the brain. The presence of myogenic tone development in the cerebral circulation has been confirmed in ex vivo studies, which showed that isolated cerebral arteries constrict or dilate in response to changes in pressure and have been suggested that myogenic reactivity is an important contributor to autoregulation of cerebral blood flow. ⁶³ Such mechanisms may extend beyond the intracranial vessels as large (extracranial) vessels also play a role in controlling blood flow resistance to the cerebral arteries. ⁵³ Thus, the increased stiffness and reduced myogenic tone development in HCD postpartum rats compared with CD animals may also help explain the increased burden of cardiovascular disease (such as stroke) after preeclampsia. ²

In summary, our study demonstrated that excessive hypercholesterolemia in pregnancy has long‐term implications for the maternal vascular system with varying impacts across different vascular beds. In fact, the larger conduit arteries (ie, carotid arteries and aortas), rather than the smaller resistance‐sized vessels (ie, mesenteric and coronary arteries), were the most impacted later in life by exposure to excessively high cholesterol levels in pregnancy. It is important to highlight that the larger conduit arteries are also the most susceptible for the development of atherosclerosis; however, we have not observed signs of subclinical atherosclerosis (ie, increased carotid intima media thickness) between the CD and HCD postpartum groups. This agrees with our previous study, in which we also did not observe changes in lipid deposition and CD68 positive cell infiltration into aortas from HCD postpartum mice. ²⁸ Thus, additional studies using transgenic models would help to dissect the link between excessive hypercholesterolemia in pregnancy and maternal atherosclerosis development later in life, as wild‐type mice and rats lack the cholesteryl ester transfer protein, which confers protection against the development of atherosclerosis. ⁶⁴ Nevertheless, our data further our understanding of the mechanisms underlying later‐life maternal vascular dysfunction after a pregnancy complicated by excessive hypercholesterolemia in pregnancy, leading to preeclampsia.

Sources of Funding

This work was supported by a foundation grant from the Canadian Institutes of Health Research (CIHR FS154313) and by the Women and Children's Health Research Institute through the generosity of the Stollery Children's Hospital Foundation and the Alberta Women's Health Foundation. Drs de Oliveira and Graton were supported by a Women and Children's Health Research Institute postdoctoral fellowship through the generosity of the Stollery Children's Hospital Foundation and the Alberta Women's Health Foundation. Dr Davidge is a Distinguished University Professor at the University of Alberta. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Disclosures

None.

Supporting information

Table S1

Figures S1–S6