Eplerenone lowers maternal blood pressure in a model of leptin-induced preeclampsia, but decreases fetal growth when administered mid-, but not late-, gestation

Abstract

Preeclampsia induces adverse cardiovascular outcomes for both mother and offspring. We established a novel leptin-induced mouse model of preeclampsia that induces hypertension, endothelial dysfunction, and fetal growth restriction, which are collectively ablated by endothelial cell mineralocorticoid receptor (MR) deletion. However, literature lacks preclinical evidence to use MR antagonism for preeclamptic patients. We hypothesize eplerenone improves blood pressure, vascular function, and fetal outcomes in leptin-infused pregnant mice. We infused timed-pregnant Balb/c mice with saline (sham) or leptin via s.c. osmotic minipump and administered vehicle or eplerenone from gestation day (GD)11-18 and GD15-18. We measured mean arterial blood pressure (BP) via radiotelemetry, vascular function in 2^nd order mesenteric arteries by wire myography, and pup/placental weights on GD18. Eplerenone from GD11-18 ablated leptin-induced increases in BP but independently decreased fetal weight and placental efficiency. Eplerenone increased vascular contractility to phenylephrine and increased mRNA expression of NADPH oxidase (NOX) 1 and 2 in the placentas of pregnant mice in the GD11-18 cohort. We observed in our GD15-18 cohort that eplerenone no longer decreased fetal weight nor placental efficiency and there was no increase in contractility to phenylephrine. In conclusion, our data suggest that although eplerenone improves leptin-induced hypertension in pregnant mice, eplerenone reduces fetal weight when administered at mid-, but not late-, gestation in pregnant mice.

New & Noteworthy

There are limited advances in treatment for preeclampsia. Leptin induces preeclampsia dependent on mineralocorticoid receptor (MR) activation, however, there is little preclinical data on the use of MR antagonists in hypertensive pregnancy. When administered at midgestation in hypertensive mouse pregnancy, eplerenone lowers blood pressure, but increases vasoconstriction in mesenteric arteries and reduces fetal growth. When administered later in pregnancy, eplerenone no longer restricts fetal growth nor increases vasoconstriction.

Introduction

Preeclampsia is a pregnancy hypertensive disorder that occurs after the 20^th week of gestation with the presentation of elevated blood pressure and a symptom of end-organ damage that increases the risk for preterm labor and fetal growth restriction (1-3). This pregnancy syndrome affects 5-8% of pregnancies worldwide (1, 3, 4) yet there is no cure and few treatments are available. Treatment strategies currently focus on ameliorating the hypertension and prolonging the pregnancy (1-4).

Leptin is a critical growth factor for a healthy pregnancy (5, 6) but inappropriately high levels of this adipokine are associated with preeclampsia, particularly severe preeclampsia (5-7). We established a mouse model of preeclampsia in which we infuse leptin from mid-late gestation, which we showed induces key characteristics of preeclampsia including elevated blood pressure and reduced fetal growth (8). We further showed that the ability of leptin to induce clinical characteristics of preeclampsia in mice requires expression of the endothelial cell mineralocorticoid receptor (ECMR) (8, 9). Current therapies to treat hypertension in pregnancy often utilize calcium channel and ß blockers (10) due to teratogenic effects on the fetus using other antihypertensive drugs like angiotensin ll receptor blockers and angiotensin-converting enzyme inhibitors. However, MR antagonist use in leptin-induced preeclampsia in preclinical models is unexplored.

There have been limited case studies on using MR antagonism in pregnancy, mainly in cases of resistant hypertension (11-14). There are two approved steroidal MR antagonists in the US, spironolactone and eplerenone. The older drug, spironolactone, is not highly specific for the MR and due to its non-specificity and affinity for androgen receptors, is contraindicated in pregnancy (11). Eplerenone is a highly specific MR antagonist, and although not approved for use in pregnancy by the FDA, has been used in several case reports for the treatment of resistant hypertension (12, 13) and primary aldosteronism in pregnancy (14-16). However, there is little preclinical data on the use of eplerenone for the treatment of preeclampsia, therefore, we sought in this study to test the hypothesis that eplerenone treatment will ablate leptin-induced preeclampsia without adverse outcomes to fetal growth in a mouse model.

Materials and Methods

Experimental Animals

All protocols were approved by the Institutional Animal Care and Use Committee of Augusta University (IACUC Protocol # 2011-0108). Mice were kept in double housing at room temperature (20-25°C), in humidity-controlled rooms (30-70%), on a 12h light, 12h dark cycle while having free access to standard chow and water. Female Balb/c mice (9-10 weeks of age) were utilized for timed pregnancies with gestation day (GD) 1 recorded by the presence of vaginal plug. Ultrasound imaging (GD11) and weight tracking were used to confirm the progression of pregnancy. Pregnant mice not on telemetry were housed in groups (2/cage) while, due to requirements for PAC-10 telemeter recording, all telemetry animals were recorded singly housed. All experimental groups were randomly assigned following successful mating and confirmation of pregnancy at GD11, after telemetry implantation. Therefore, mating efficiency and telemetry implantation surgery failure did not contribute to varying N values during pregnancy (only for the prepregnancy recordings) and this insured that all groups were done alternatively. Varying N values in our non-telemetry data is due to pregnancy loss (of which we experienced equally among groups), failure of hanged mesenteric vasculature to meet criteria for analysis (failure of vessels to normalize or respond to KCl in initial part of protocol) and limitation of sample availability of plasma or placenta tissue. For telemetry experiments, varying N values are primarily due to failure of telemeter recording during pregnancy. Telemeter recording was discontinued prior to breeding and resumed on GD11, in many animals this span of time resulted in a clotted or displaced telemeter which was determined based on Ponemah software thresholds for variability, or a non-physiological recording (i.e. mean arterial pressure of <70 in a conscious animal). As data was analyzed simultaneously after recording in order to provide blinding and prevent bias, these failures were not discovered until the experiment conclusion. Animals were euthanized in the morning (the sleep/light cycle of the mouse) of GD18 with 5% isoflurane anesthesia followed by thoracotomy. Fetuses were euthanized by decapitation. Maternal weight was reported as the weight of the dam on GD18 with the total weights of the fetus/placenta subtracted.

Leptin Infusion

On GD11, mice were randomly assigned to either sham or leptin infusion as previously described (8, 9). Briefly, mice were anesthetized with 1.7% isoflurane with body temperature maintained at 37°C using a heating pad and a dorsal incision made after sterilizing the incision area. Then, a mini osmotic pump (ALZET, Cupertino, CA, USA; model 1007 D, 0.5μL/hs, 7 day) containing leptin (CYT-351, Prospec, Rehovot, Israel, 0.9mg/kg/day) was implanted subcutaneously after creation of a pocket with sterile saline while sham surgery consisted of similar procedure with a saline bolus in place of osmotic pump (8, 9). Dorsal incisions were closed with 2 staples and slow-acting (72 hour release) buprenorphine (0.6-1.2mg/kg) was administered subcutaneously as an analgesic prior to surgery.

Eplerenone Administration

Two different administration times were utilized for eplerenone treatment: GD11-18 and GD15-18. At first, GD11-18 was used in tandem with leptin treatment. However, once independent fetal growth restriction was observed, a later administration of eplerenone (GD15-18) was implemented to see if fetal growth would still be affected. This specific timepoint was chosen due to knowledge that plasma volume expansion peaks after the halfway point in a pregnancy (17) and seeking to administer the drug after this peak. For both times, mice were randomly assigned to either vehicle (ORA-Plus and ORA-Sweet, Cat: 574030316/574030216, McKesson, Irving, TX, USA) or eplerenone (29057-5G, Chem-Impex, Wood Dale, IL, USA, 200mg/kg/day) treatment in their drinking water (18). Calculations for eplerenone solution were with the assumption that mice drink on average 4mL a day based on metabolic data obtained in our laboratory in pregnant mice.

Conscious Blood Pressure Measurements

For blood pressure measurement, a specific cohort of mice was implanted with radio telemeters (PA-C10, Data Sciences International (DSI), St. Paul, MN, USA) in the left carotid artery via catheter for continuous measurement under 2.5% isoflurane. Animals were allowed 7 days to recover before recording baseline pressures and heart rate measurements for 5 days. Breeding then occurred with GD15-18 measurements recorded and analyzed as described by our group and others (8, 9, 18). n values vary between groups and between baseline and pregnancy due to telemeter failure in the pregnancy recordings.

Vascular Reactivity

Second order mesenteric resistance arteries were identified, cleaned of perivascular adipose tissue, isolated 2mm rings, and mounted on wire myography (DMT, Ann Harbor, MI, USA) as described (19) for assessment of vascular function. Each animal had 4 rings isolated and then the mean of the measures was calculated from these rings. Vessels were normalized using the DMT normalization module and left to equilibrate for 20 minutes before being subjected to KCl (80mM) for maximum depolarization contractions. There were no differences in maximal KCl constriction in any groups. Constriction response to Phenylephrine (Phe, 1nM-100uM) were performed and expressed as a percentage to KCl. Phe (30μM) was used to constrict the vessels before being assessed for endothelial-independent relaxation to sodium nitroprusside (SNP) (log curves 1nM – 30μM concentrations) and responses expressed as percentage to preconstriction. Curves that were included in analysis had to pass the following rigor tests 1. The ability of the vessel to hold vascular tone (not observe a sloped decrease in tone after addition of preconstrictor or immediate loss of tone on 1^st dose), 2. To react with vasoconstriction to KCl depolarization and 3. The ability of the vessel to be normalized to baseline tension (13.1kPa). Vessel rings that did not fulfill these requirements were excluded from analysis. Data were acquired, recorded, and analyzed using LabChart software (ADInstruments, Colorado Springs, CO, USA) (19, 20).

Fetal Assessment

The mice from all cohorts were weighed before being euthanized on GD18 under isoflurane anesthesia and tissues and blood harvested. Whole blood was collected from the animal in a sterile tube containing heparinized sodium, plasma was isolated from blood by centrifugation at 12,000 relative centrifugal force for 10min. Fetuses were removed and weighed from each uterine horn as previously described (9) and their corresponding placenta extracted and weighed. Fetal demise was recorded as a percentage from total litter size. Pup and placenta total weight was subtracted from maternal weight for an adjusted weight to use for maternal measurements.

Metabolic Cage Measurements and Urine Analysis

Pregnant mice were individually placed in metabolic cages for 24hr urine collection and measurement of water and food intake. We performed metabolic cage measures at two timepoints, at GD17-18 for our cohort administered leptin/eplerenone from GD11-18 and at GD15-16 for our cohort administered leptin/eplerenone from GD15-18. Our rationale was to first assess urine volume in response to eplerenone at time of delivery and in our GD15-18 groups these measures were designed to assess initial urine volume response to eplerenone at first dosage. Urine collected from both cohorts was used for two assays: Bicinchoninic Acid (BCA, Cat: 23227, Thermo Scientific, Waltham, MA, USA) protein and thiobarbituric acid reactive substances (TBARS, Cat: 10009055, Cayman Chemical, Ann Harbor, MI, USA).

Plasma Analysis and Placental Quantitative PCR Analysis

Plasma endothelin-1 (ET-1, Cat: DET100, R&D Systems, Minneapolis, MN, USA) was quantified by commercially available enzyme-linked immunosorbent assays (ELISA) per manufacturer’s instructions. Whole placental tissue (50mg) was homogenized in TRIzol and RNA isolated with mini kit (Cat: 12183018A, Invitrogen, Waltham, MA, USA). cDNA was synthesized with reverse transcription (ThermoFisher, Waltham, MA, USA), qRT-PCR was performed with SyBr Green (Applied Biosystems, Waltham, MA, USA), targeting pre-pro endothelin-1 (PPET-1), endothelin-1 converting enzyme 1 (ECE-1), nicotinamide adenine dinucleotide phosphate oxidase (NOX) 1, 2, 4, and cluster of differentiation 31 (CD31). Primer sequences (Integrated DNA Technologies, Coralville, IA, USA) are listed in Table 1. Ct values were normalized to 18s expression (ΔCt) for housekeeping due to high expression and low variability between groups, followed by normalization to control groups (ΔΔCt) and calculation of relative gene expression (2^ΔΔCt) as published previously (8) and in accordance with Applied Biosystems guidelines for expression analysis (Applied Biosystems User Bulletin, no. 2, 1997).

Table 1.

Primers for qPCR

Target	Gene Name	Forward Sequence	Reverse Sequence
18s	Rn18s	TCGAGGCCCTGTAATTGGAA	CCCTCCAATGGATCCTCGTT
PPET-1	Edn1	GCCACAGACCAGG CAGTTAGA	CACCAGCTGCTGATAGATACACTTC
ECE-1	Ece1	CACAGCAGTTCCTGATCTCTAC	CTTGGTGGGCGAGTAGTAAG
NOX1	Nox1	CAGTTATTCATATCATTGCACACCTATTT	CAGAAGCGAGAGATCCATCCA
NOX2	Cybb	CAAGATGGAGGTGGGACAGT	GCTTATCACAGCCACAAGCA
NOX4	Nox4	TGTTGCATGTTTCAGGTGGT	AAAACCCTCGAGGCAAAGAT
CD31	PECAM1	TGCGGTGGTTGTCATTGG	TGTTTGGCCTTGGCTTTCC

Pre-pro endothelin-1 (PPET-1), endothelin-1 converting enzyme (ECE-1), NADPH oxidase (NOX) 1, 2, and 4, cluster of differentiation 31 (CD31).

Western Blot Analysis

Whole mouse placenta lysates were prepared, and proteins separated by SDS-PAGE (Bio-Rad). Following separation, proteins were transferred to polyvinylidene difluoride (PVDF) membranes (Immun-Blot, Bio-Rad). Membranes were then blocked at room temperature (22 °C) in Tris-buffered saline containing 0.05% Tween-20 (TTBS) and 5% (w/v) non-fat dry milk. Blocked membranes were incubated overnight at 4 °C with primary antibodies diluted in blocking buffer. The primary antibodies that were used are listed in Table 2. Membranes were washed and incubated for 1 hour at room temperature with horseradish peroxidase (HRP)-conjugated secondary antibodies diluted 1:1000 in blocking buffer anti-mouse IgG-HRP (Cell Signaling Technology, Cat# 7076). Protein bands were visualized using enhanced chemiluminescence (ECL) substrate (Cytiva, Cat# RPN2236) and detected using the FluorChem E imaging system (ProteinSimple). Molecular weight markers (Bio-Rad) were run in parallel to confirm the identity and size of the detected bands. Protein loading was normalized by probing for heat shock protein 90 (HSP90; BD Transduction Laboratories, Cat# 610419) which did not differ across groups but ponceau was also run to confirm loading (Supplementary Fig. 12). Briefly, before blocking, membranes were incubated with Ponceau S Staining solution (0.1% Ponceau S in 5% acetic acid, Sigma, Cat# P7170-1L) for 10min and then washed with distilled water. Membranes were then imaged, and then blocked overnight as stated above. Data were quantified using Image J software. Whole blots for all westerns blots are in Supplementary Fig. 9-11.

Table 2.

Antibodies utilized for Western Blots

Antibody	Full Name	Dilution	Source	Category #
p-eNOS (Ser1177)	Phosphorylated endothelial nitric-oxide synthase	1:1000	BD Transduction Laboratories	612393
t-eNOS	Total endothelial nitric-oxide synthase	1:1000	BD Transduction Laboratories	610297
p-ERK	Phosphorylated extracellular signal-regulated kinase	1:1000	Cell Signaling	4370
t-ERK	Total extracellular signal-regulated kinase	1:1000	Cell Signaling	4695
SOD2	Superoxide dismutase 2, mitochondrial	1:50,000	Protein Tech	24127-1-AP
GPx4	Glutathione peroxidase 4	1:1000	Protein Tech	67763-1-Ig
HSP90	Heat shock protein 90	1:1000	BD Transduction Laboratories	610419

Statistical Analysis

All data were expressed as means ± standard error (SE) and analyzed in GraphPad Prism 10 (GraphPad Software, La Jolla, CA, USA). All data was tested for normality with Shapiro-Wilk test and data transformed if needed for statistical analysis. Groups in which normality failed are mentioned in the figure legends. To measure non-repeated variables, eplerenone and leptin, two-way ANOVA was run with Tukey’s post hoc test for multiple comparisons. Myography analysis was performed with two-way ANOVA with repeated variable measures. For all comparisons, P values <0.05 were considered significant. ROUT test was run to identify any outliers and values excluded from final analysis.

Results

Eplerenone increases heart weight, but not water intake or urine excretion, in leptin-treated pregnant dams

In mice treated with leptin +/− eplerenone from GD-11-18, dams were euthanized and their body, kidney, and heart weights measured. Body weight measured on GD18 was not different between dams of all groups (Fig. 1A) (8). Neither leptin nor eplerenone increased kidney to body weight ratios (Fig. 1B), however, eplerenone increased heart to body weight ratios (Fig. 1C) primarily driven by the leptin+eplerenone group. Dams were placed in metabolic cages from GD17-18 to record 24hr water and food intake and urine excretion volume. Neither leptin nor eplerenone significantly affected urine excretion volume/day, water or food intake (Fig. 1D-F). We measured protein in the urine collected from the metabolic cages using a BCA kit, as proteinuria is a diagnostic marker in the clinic with patients. We did not see any difference in any groups (Fig. 1G).

Figure 1. Dams on eplerenone from GD11-18 have greater heart weight ratios but no effect on 24hr urine volume.

Dam (A) body weight (B) kidney to body weight ratio, and (C) heart to body weight ratio on GD18 adjusted to total pup and placenta weight. Metabolic cage 24hr data collection of (D) food intake (E) urine output, (F) water intake, and (G) protein concentration in urine measured by BCA kit adjusted to 24hr volume excreted. Means ± SE, 2Way ANOVA with Tukey’s post hoc for multiple comparisons. Data in panel B were normalized prior to analysis, all other data passed normality. ROUT test excluded an outlier in panel B and F. n= 5-17. *p<0.05, **p<0.01.

Eplerenone decreases blood pressure but promotes vascular contractility in leptin-infused pregnant mice

Conscious blood pressure and heart rate (HR) were recorded before and during pregnancy in our GD11-18 cohort. Mean arterial pressure (MAP), systolic blood pressure (SBP), diastolic blood pressure (DBP) and HR did not differ between treatment groups before pregnancy (Supplementary Fig. 1). Similar to our published results (8), leptin infusion increased MAP and SBP in pregnant mice GD15-18 compared to sham vehicle (Fig. 2A,B), and trended to increase DBP (p=0.1075) and HR (p=0.0579) (Fig. 2C,D). Sham+eplerenone also increased MAP and SBP (Fig. 2A,B) from GD15-18. However, combination of leptin+eplerenone ablated the ability of leptin or eplerenone to increase MAP or SBP shown by a lack of statistical difference from leptin+vehicle and a persisting interaction effect of leptin and eplerenone. On GD18, 2^nd order mesenteric vessels were assessed for vascular function. Vascular relaxation to SNP did not differ between treatment groups (Fig. 2E). Eplerenone increased maximal constriction to Phe both in the absence and presence of leptin measured both as 2-Way ANOVA of concentration response curves (Fig. 2F) and at the maximal dose, E_max (Fig. 2G). There was no difference in maximal KCl constriction (Fig. 2H).

Figure 2. Eplerenone decreased leptin-induced hypertension.

Blood pressure of dams measured as mean of GD15-18 as shown by (A) mean arterial pressure (MAP), (B) systolic, and (C) diastolic blood pressure (SBP, DBP). (D) Heart rate of pregnant mice from GD15-18. Vascular relaxation responses to (E) sodium nitroprusside (SNP). (F) Vascular constriction to phenylephrine (Phe) with (G) E_max and (H) constriction to KCl. Means ± SE, 2Way ANOVA with Tukey’s post hoc for multiple comparisons and repeated measures on E-F. Telemetry: n= 3-6; Myography: n= 9-11. *p<0.05

Both leptin and eplerenone induce fetal growth restriction

We reported previously that leptin decreases fetal growth (8). In this report, eplerenone decreased fetal weight and a significant eplerenone-by-leptin interaction showed this effect persisted across leptin groups (Fig. 3A). However, these effects were muted when averaging by dam (Supplementary Fig. 2A). Leptin increased placenta weight compared to sham vehicle and there persisted a significant interaction of eplerenone-to-leptin for placenta weight across leptin groups (Fig. 3B) and this effect persisted in dam average of placenta weight (Supplementary Fig. 2B). Eplerenone and leptin decreased placental efficiency (Fig. 3C). Neither leptin nor eplerenone impacted litter size (Fig. 3D) nor increased fetal death as shown as a percentage of reabsorptions to total litter size (Fig. 3E). To measure vascularity in the placentas, we measured CD31 mRNA expression but did not see a difference between any groups (Supplementary Fig. 3A).

Figure 3. Eplerenone GD11-18 induces fetal growth restriction.

(A) Pup weight, (B) placenta weight and (C) placental efficiency as measurement of fetal growth taken GD18 as well as (D) litter size. (E) Fetal death by count of reabsorptions per litter shown as a percentage. Means ± SE, 2Way ANOVA with Tukey’s post hoc for multiple comparisons. All panels were normalized prior to analysis. n= 10-18 litters. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001

Eplerenone does not decrease endothelin-1 plasma concentration in leptin-infused pregnant mice but increases NOX1&2

Our previous report (8) showed that leptin increased placental gene expressions of ECE-1 and PPET-1. In our GD11-18 cohort, we show that leptin increased plasma ET-1, PPET-1, and ECE-1 (Fig. 4A-C). Our previous report also showed that deletion of ECMR decreased ECE-1 and PPET-1 mRNA expression, however, eplerenone from GD11-18 did not decrease plasma ET-1, nor placental PPET-1 and ECE-1 mRNA in leptin-treated mice. We further measured NOX1, 2, and 4 expressions in placenta and observed that leptin alone had no effect to increase NOX1 or 4 expressions. However, the main effect of eplerenone increased NOX1 which persisted with the interaction of leptin+eplerenone in NOX1 and 2, but not 4, in placenta (Fig. 4D-F). There was no difference in p-eNOS to total eNOS ratio across any of our groups, however there was an increase in total eNOS from the main effect of eplerenone (Supplementary Fig. 4). We measured lipid peroxidation with TBARS assay, but there was no difference between any groups (Supplementary Fig. 5A). We also measured the protein expression of the phosphorylation of ERK (pERK) to total ERK (tERK), SOD2, and GPx4. Despite the differences seen in NOX 1 and 2 mRNA expression, there was no difference between groups in these proteins (Supplementary Fig. 6).

Figure 4. Eplerenone treatment GD11-18 does not decrease ET-1 in pregnancy.

(A) Plasma endothelin-1 (ET-1) secretion. (B) Pre-pro endothelin 1 (ppET-1), (C) endothelin converting enzyme 1 (ECE-1), (C) NADPH oxidase 1 (NOX1) and (D) NOX2, and (E) NOX4 mRNA expression in mouse placenta. Means ± SE, 2Way ANOVA with Tukey’s post hoc for multiple comparisons. Data in panel F normalized before analysis, all other data passed normality. ROUT test excluded an outlier in panel E. n= 6-13. *p<0.05, **p<0.01, ***p<0.001

Eplerenone administered in late gestation does not promote increased heart weight or increases in urine volume

We repeated leptin infusion from GD11-18, with eplerenone treatment only late in gestation, GD15-18. This GD15-18 cohort was decided after seeing the independent effect eplerenone had to decrease fetal weight when administered GD11-18. Similar to what we observed in the GD11-18 cohort, eplerenone had no effect on dams’ body weight (Fig. 5A). Neither eplerenone nor leptin affected kidney or heart to body weight ratio (Fig. 5C,D). Pregnant mice given eplerenone from GD15-18 were placed in metabolic cages for the first 24hrs of eplerenone administration, to determine the early effects of eplerenone on potential diuresis, and their water and food intake measured as well as urine excretion volume. There was no change in urine excretion or food intake across groups at GD15-16 (Fig. 5D,F). Interestingly, in this cohort of mice, the main effect of leptin did decrease water intake, particularly driven by the leptin+eplerenone group, which demonstrated a reduced water intake compared to both leptin+vehicle and sham+eplerenone groups (Fig. 5E). When we measured protein concentration in the urine collected from the GD15-18 cohort, similar to the GD11-18 cohort, there was no difference across groups (Fig.5G).

Figure 5. Dams on eplerenone from GD15-18 experience no heart weight to body weight ratio increase and have no effect on 24hr urine output.

Dam (A) body weight, (B) kidney to body weight ratio, and (C) heart to body weight ratio on GD18 adjusted to total pup and placenta weight. Metabolic cage 24hr data collection of (D) food intake, (E) urine output, (F) water intake, and (G) protein concentration in urine measured by BCA kit adjusted to 24hr volume excreted. Means ± SE, 2Way ANOVA with Tukey’s post hoc for multiple comparisons. Data in panel D were normalized prior to analysis, all other data passed normality. ROUT test excluded an outlier in panel B and F. n= 4-18. *p<0.05, **p<0.01

Late gestation eplerenone treatment does not increase vascular contractility nor decrease fetal growth in pregnant dams

Similar to our GD11-18 cohort, and compared to vehicle sham and leptin mice, we observed no changes in SNP-mediated relaxation in mesenteric arteries of pregnant dams in response to eplerenone GD15-18 (Fig. 6A). Eplerenone GD15-18 trended to increase constriction overall (sham vehicle vs sham eplerenone p=0.0578), significantly shown in the E_max (Fig. 6B,C), but there was no difference in KCl maximal constriction (Fig. 6D). In contrast to the GD11-18 eplerenone, GD15-18 administration of eplerenone did not decrease pup weight, placenta weight, nor placental efficiency (Fig. 6E-G), also reflected in dam averages per litter (Supplementary Fig. 7A,B). Interestingly, GD15-18 eplerenone did decrease litter size and fetal reabsorptions (Fig. 6H, I) contrary to what we observed in the eplerenone GD11-18 cohort. Later administration of eplerenone did not impact CD31 mRNA expression, similar to what was observed in the GD11-18 cohort (Supplementary Fig. 3B).

Figure 6. Eplerenone increased pup weights when administered GD15-18.

Vascular relaxation responses to (A) SNP, (B) vascular constriction to Phe with (C) E_max and (D) constriction to KCl. (E) Pup weight, (F) placental weight, (G) placental efficiency, (H) pup count, and (I) fetal death when eplerenone is administered GD15-18. Means ± SE, 2Way ANOVA with Tukey’s post hoc for multiple comparisons. Data in panel D-F, H and I were normalized prior to analysis, all other data passed normality. ROUT test excluded an outlier in panel F and I. n= 4-6 dams (vascular) or litters (pup data). *p<0.05, **p<0.01

Late gestation administration of eplerenone decreased ET-1 levels in leptin-treated pregnant dams

Eplerenone treatment from GD15-18 decreased ET-1 plasma level and placental PPET-1 but not ECE-1 mRNA expressions in mouse placenta (Fig. 7A-C), in contrast to the GD11-18 cohort. Similar to the GD11-18 cohort, GD15-18 eplerenone treatment increased NOX1&2 but not 4, mRNA expression in mouse placenta in the presence of leptin (Fig. 7D-F). We also measured lipid peroxidation in this cohort, but similarly, there was no difference between any groups (Supplementary Fig. 5B). Interestingly, in our eplerenone groups administered GD15-18, there was a significant increase in pERK/tERK protein expression (Supplementary Fig. 8A) but no difference in SOD2 or GPx4 protein expression (Supplementary Fig. 8B, C).

Figure 7. Eplerenone administered GD15-18 does not affect ET-1 production.

(A) Plasma endothelin-1 (ET-1) secretion. (B) ppET-1, (C) ECE-1, (D) NOX1, (E) NOX2, and (F) NOX4 mRNA expression in mouse placenta from GD15-18 cohort. Means ± SE, 2Way ANOVA with Tukey’s post hoc for multiple comparisons. All data passed normality. n= 5-6. *p<0.05

Discussion

In the current study, we tested the hypothesis that eplerenone treatment would ablate key characteristics of preeclampsia in pregnant mice that are infused with exogenous leptin. In line with our expectations, we observed that eplerenone reduced leptin-induced increases in blood pressure when administered from mid-late gestation (GD11-18). However, we observed strikingly that eplerenone reduced fetal weight independently of leptin in these pregnant mice. We observed that this decrease in fetal weight was associated with an effect of eplerenone to increase vascular contractility to phenylephrine in mesenteric resistance arteries as well as increase expression of NOXs 1 and 2 in the placentas of these mice. Following our collection of these data we sought to determine whether the timing (mid-gestation) of eplerenone treatment was a candidate culprit leading to reduced fetal growth. To address this question, we administered eplerenone only in late gestation from GD15-18 and we observed that effects of eplerenone to reduce fetal weight were ablated. These data indicate that eplerenone treatment in hypertensive pregnant mice is least likely to harm fetal growth in late gestation, but could promote negative outcomes when given in mid-gestation.

In this study, we reproduce our previous data showing that leptin decreases fetal growth and placental efficiency (8, 9). In our 2022 report (8), we showed that endothelial MR deletion protected leptin-infused mice from fetal growth restriction, therefore, we initially hypothesized that eplerenone would improve fetal growth in leptin-infused pregnant mice. Contrary to our hypothesis, eplerenone in the GD11-18 treatment group decreased fetal weight, litter size, and placental efficiency. During pregnancy, peak plasma volume expansion occurs in roughly mid-gestation (17), around GD9 to GD13 in a mouse pregnancy (21, 22). We did not, however, observe changes to water intake or urine volume at GD17-18 in this cohort. We, therefore, made a second hypothesis that the timing of eplerenone at the mid-gestation timepoint may have enacted a diuresis effect, impacting fetal growth, that was no longer seen at the GD17-18 point where we collected urine and measured water intake. We observed that GD15-18 eplerenone did ablate the reduction of fetal growth induced by eplerenone in the GD11-18 cohort. We also no longer saw a decrease in placental efficiency and no effect of eplerenone on litter size. We measured water intake/urine excretion in the 1^st 24 hours in our GD15-18 cohort and eplerenone did not have an apparent diuresis effect. Therefore, either eplerenone has a diuresis effect at mid-gestation that it does not have when given at GD15 or another mechanism is likely at play that reduced fetal growth in our GD11-18 cohort.

We speculate there may be oxidative stress generated in the placenta by eplerenone in our pregnant mice. When measuring mRNA expression of whole mouse placenta, NOX1 and 2 were significantly increased by eplerenone. Zerrad-Saadi and others have shown the importance of NOX in superoxide generation in the placenta and that NOX complexes are upregulated in preeclampsia (23-27). This increase in NOX 1 and 2, regardless of the timing of administration of eplerenone, indicates there is increased ROS generation in the mouse placenta in our eplerenone treated dams that may contribute to reduced placental perfusion in mid-gestation. To test resolution of this potential ROS, we measured SOD2 and GPx4. If we saw an increase in these proteins, we may assume a resolution of ROS generation. However, there was no difference in either cohort between any groups. We also measured pERK to tERK as an indirect measurement of cell stress. Though there was no difference in our GD11-18 cohort, in the GD15-18 cohort there was a significant increase in protein expression in the eplerenone groups. These data indicate a need for further investigation into direct measures of cell stress and ROS generation. It also has been shown that increases in NOX can lead to hypertension and vasoconstriction (28, 29), potentially due to increases in ROS.

We have shown previously that leptin increases blood pressure in pregnant mice (8) and that knockout of endothelial MR ablates this hypertension (8, 9). Our data shows that eplerenone, an MR antagonist, improves hypertension in leptin-infused pregnant mice. However, in our sham+eplerenone group, there was increased blood pressure as well. We propose this may be attributable to increased vasoconstriction as shown in our vascular reactivity data. Eplerenone, in both administration cohorts, had increased constriction to phenylephrine, which may be indicative of increased sympathetic activation (30). This is counterintuitive to have an ablation in hypertension yet increased vasocontriction, indicating hyperactivity of the resistance vessels. We also show increased heart weights in eplerenone groups in the GD11-18 group, but not in the GD15-18 treatment group, which may be indicative of increased vascular resistance and cardiac hypertrophy as a result in the GD11-18 cohort. These data show that global antagonism of MR in pregnancy may increase sympathetic activation and vascular hyperactivity, leading to potential cardiac hypertrophy, reducing fetal growth in the earlier administration timepointPLR. An increase in vascular contractile response to phenylephrine coupled with a decrease in systemic blood pressure implies that other blood pressure lowering mechanisms of eplerenone, such as endothelial MR activation on the vascular endothelium, must have been activated to sustain a decrease in blood pressure.

In our previous study (8), leptin increased placental ET-1 production but only with an intact endothelial MR in pregnant mice. Others also report that ECMR activation increases ET-1 production (31). Our data shows that only later eplerenone administration decreases ET-1 production in placentas from leptin infused dams. Therefore, eplerenone treatment for a shorter period in late gestation lowered placental ET-1. We speculate that this later administration versus earlier decreased ET-1 in the presence of leptin due to overactivation of ECMR. We speculate that leptin’s overactivation of ECMR is what differentiates this dichotomy between treatment timepoints and is reflected in ET-1 levels. ET-1 is increased in preeclamptic patients (32, 33) compared to healthy patients and is believed to be a major mechanism whereby vascular resistance increases in these patients. Both ET-1 and leptin independently have shown to be increased in preeclamptic patients (32, 34-37), however our data suggest that leptin leads to increased ET-1. Increased ET-1 has been observed in other models of preeclampsia like the reduced uterine perfusion pressure model and soluble FMS-like tyrosine kinase-1 elevation model (32), showing that ET-1 potentially independently induces endothelial dysfunction that is seen in preeclampsia. Leptin also is known to impact the vasculature due to through increased ECMR activation. Our data implies that lowering ET-1 may be an early response to MR antagonism. Tying together this impact of leptin increasing ET-1 and its activation of ECMR leads to our speculation that this lack of suppression of ET-1 by eplerenone may remove a protective mechanism against vascular constriction in leptin-induced hypertensive mice.

There are limitations to this current study. We speculate that eplerenone may be increasing the generation of superoxide at the placenta, but we have no direct ROS measurements in this study, though we do measure GPx4 and SOD2. We also did not look at any fetal effects beyond weight and paired placenta weight so looking at long term fetal effects may better explain why eplerenone was limiting fetal growth when given at the midgestation timepoint but not later. We only measured blood pressure in our GD11-18 eplerenone mice groups, as these data indicated that fetal growth did not correlate with blood pressure improvement. However, we would speculate that eplerenone would decrease leptin-induced hypertension even if only given from GD15-18 due to clinical reports lowering blood pressure in a short period in pregnancy (13). In the current study, we did not identify fetal sex prior to utilizing placentas for experimental analysis. Recent studies (38, 39) have shown the importance of fetal sex programming on the placenta. Without the identification of sex, we are unable to observe the specific implications of male versus female placentas. Also, in our drug calculations for administering eplerenone, it is done with the assumption that mice drink approximately 4mL of water in a 24hr period, based on previous data from the laboratory. In our GD15-18 group, the leptin eplerenone group has a mean closer to 2mL for a 24hr period from GD15-16. While this did reach statistical significance compared to the vehicle eplerenone group, this did not result in an overall effect of eplerenone in the results nor was this group statistically significantly lower than any other group. Therefore, while there may have been some slight variation in the day to day drug administration (as is seen in humans) we conclude that our mice overall had similar drug administration.

Clinical Perspectives and Significance

In summary, our data support that eplerenone treatment reduces blood pressure in hyperleptinemia-induced hypertensive pregnant mice but may have adverse effects to reduce fetal growth when given in mid-, but not later, gestation. This study adds to the limited existing pre-clinical data of eplerenone use in pregnancy (40). There are several case reports in which eplerenone is used in pregnant women (11-16, 41-43), all in which blood pressure is elevated and the use of eplerenone decreased the patients’ blood pressure. Some patients that have been reported started eplerenone treatment before conception as part of their management of resistant hypertension (16, 43), with one developing FGR (16), but due to other complications arising (preeclampsia diagnosis), it cannot be concluded if this was an effect of treatment. Our data combined with these case reports does suggest that the blood pressure lowering effect of eplerenone is not sufficient to improve fetal outcomes in the mid-gestation period. Targeting MR receptor mechanisms may not be ideal in pregnancy as there have been studies showing that finerenone, a non-steroidal MR antagonist, negatively impacts fetal growth (44) Overall, clinical data and our current study suggest that eplerenone may be an effective blood pressure-lowering agent in hypertensive pregnancy, but caution and as late-gestation as possible treatment should be considered when administering eplerenone due to potential impacts to fetal growth.

Supplementary Material

Supplemental Figures 1-12: 10.6084/m9.figshare.30642770

Data Availability

Data will be made available upon reasonable request.