Prenatal Arsenite Exposure Alters Maternal Cardiac Remodeling During Late Pregnancy

Abstract

Exposure to inorganic arsenic through drinking water is widespread and has been linked to many chronic diseases, including cardiovascular disease. Arsenic exposure has been shown to alter hypertrophic signaling in the adult heart, as well as in utero offspring development. However, the effect of arsenic on maternal cardiac remodeling during pregnancy has not been studied. As such, there is a need to understand how environmental exposure contributes to adverse pregnancy-related cardiovascular events. This study seeks to understand the impact of trivalent inorganic arsenic exposure during gestation on maternal cardiac remodeling in late pregnancy, as well as offspring outcomes. C57BL/6J mice were exposed to 0 (control), 100 or 1000 μg/L sodium arsenite (NaAsO₂) beginning at embryonic day (E) 2.5 and continuing through E17.5. Maternal heart function and size were assessed via transthoracic echocardiography, gravimetric measurement, and histology. Transcript levels of hypertrophic markers were probed via qRT-PCR and confirmed by western blot. Offspring outcomes were assessed through echocardiography and gravimetric measurement. We found that maternal heart size was smaller and transcript levels of Esr1 (estrogen receptor alpha), Pgrmc1 (progesterone receptor membrane component 1) and Pgrmc2 (progesterone receptor membrane component 2) reduced during late pregnancy with exposure to 1000 μg/L iAs vs. non-exposed pregnant controls. Both 100 and 1000 μg/L iAs also reduced transcription of Nppa (atrial natriuretic peptide). Akt protein expression was also significantly reduced after 1000 μg/L iAs exposure in the maternal heart with no change in activating phosphorylation. This significant abrogation of maternal cardiac hypertrophy suggests that arsenic exposure during pregnancy can potentially contribute to cardiovascular disease. Taken together, our findings further underscore the importance of reducing arsenic exposure during pregnancy and indicate that more research is needed to assess the impact of arsenic and other environmental exposures on the maternal heart and adverse pregnancy events.

INTRODUCTION

Inorganic arsenic (iAs) is a widespread drinking water contaminant which ranks as the top chemical on the Agency for Toxic Substances and Disease Registry’s (ATSDR) Substance Priority List¹. iAs is naturally found in the Earth’s crust, and activities such as mineral extraction, waste processing, pesticide application, and additives to poultry and swine feed have mobilized iAs into food and drinking water supplies². While also found in air, the greatest public health threat stems from the ingestion of iAs-contaminated groundwater, and food prepared with such groundwater^2,3. There are two forms of iAs, trivalent and pentavalent arsenic, with trivalent arsenic being the most toxic form⁴. iAs has been linked to the etiology of many chronic diseases, including cancer, chronic respiratory disease, and many forms of cardiovascular disease (CVD)^5–13, including hypertension, atherosclerosis, coronary heart disease, and stroke¹⁴. Research from our group demonstrated that chronic iAs exposure induced sex-specific pathological remodeling of the adult heart and altered susceptibility to ischemic injury^15,16. However, studies have yet to examine the impact of iAs on the cardiovascular system during physiological stressors, such as pregnancy.

During pregnancy, the heart undergoes hypertrophic remodeling, increasing chamber size and wall thickness in response to the increase in blood volume that occurs during pregnancy^17,18. Blood volume in the pregnant mother increases significantly starting at 6 weeks of gestation, resulting in a 45% increase in blood volume compared to non-pregnant women^19–21. Cardiac output also increases during pregnancy and remains constant after the end of the second trimester^22,23. Pregnancy-induced hypertrophy is reversible^24,25 and regulated by various phases of hormonal signaling during pregnancy. Both progesterone and estradiol increase throughout pregnancy and have been shown to modulate cardiac hypertrophy^26,27. Progesterone is known to induce cardiomyocyte hypertrophy, while estradiol has a biphasic effect, promoting hypertrophy at low concentrations and inhibiting hypertrophy at higher concentrations^28,29.

Emerging evidence suggests that the environment plays a role in cardiovascular-related pregnancy complications. Exposure to air pollution and metals is associated with reduced conception rates, infertility, and increased pregnancy loss^30–32, suggesting interference with hormone signaling. However, it is unclear how exposure to environmental factors is associated with peripartum cardiomyopathy and other cardiovascular complications during pregnancy, as the mechanism through which the environment impacts the cardiovascular system during pregnancy has yet to be explored. Metals, such as iAs, have negative effects on the cardiovascular system, which undergoes physiological alterations during pregnancy^8,33–36. However, the pathways and physiologic processes that are altered with iAs exposure during pregnancy have not been established. In the US, 23.8 maternal deaths occurred for every 100,000 live births in 2020 and CVD is the leading cause of pregnancy-related mortality^37,38. Women who develop complications during pregnancy, such as preeclampsia, hypertension, or gestational diabetes, are also at higher risk for adverse pregnancy outcomes and CVD postpartum³⁹.

Our previous studies demonstrated that iAs exposure altered cardiac hypertrophy in adult male mice, and we next sought to examine the impact of iAs exposure on maternal cardiac remodeling during pregnancy^15,16. Pregnant C57BL6/J mice were exposed to either 0, 100 or 1000 μg/L iAs in drinking water during gestation, and a late pregnancy (E17.5) endpoint was examined. We found that pregnant female mice exposed to 1000 μg/L iAs showed blunted maternal cardiac growth and decreased transcriptional levels of estrogen receptor alpha (ERα), progesterone receptor membrane component 1 and 2 (Pgrmc1/2), and atrial natriuretic peptide (ANP), and decreased protein expression of Akt in the maternal heart during late pregnancy.

METHODS

Animals and Arsenic Exposure

This investigation conforms to the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health (NIH; Publication No. 85–23, Revised 2011) and was approved by the Institutional Animal Care and Use Committee of Johns Hopkins University. Timed pregnant female C57BL/6J mice were purchased from Jackson Laboratory (Bar Harbor, ME). Mice (2 per cage) were housed under specific pathogen-free conditions and maintained on AIN-93G chow (Research Diets, New Brunswick, NJ) and Pure Life water (Nestle Waters North America, Stamford, CT). Research Diets has reported levels of iAs in AIN-93G chow to be below the limit of detection, and our lab has previously confirmed levels of iAs in Pure Life water to be below the limit of detection using inductively coupled plasma mass spectrometry measurements^15,16. Pregnant and non-pregnant female mice (6–12 weeks of age) were given Pure Life water containing 0 (control), 100 or 1000 μg/L sodium arsenite (NaAsO₂) beginning at embryonic (E) 2.5 and continuing until harvest at E17.5 or parturition (Fig. 1). Fresh water bottles containing control or NaAsO₂ drinking water were replenished every 2–3 days to maintain concentrations of iAs and minimize oxidation. Body weight, and food and water consumption did not significantly differ between treatment groups throughout the exposure (Supplemental Figure 1).

Figure 1: Diagram of Experimental Timeline and Endpoints.

Experimental timeline of iAs exposure and experimental endpoints for dams and offspring. Created with Biorender.

Tissue Harvest and Collection

Mice were anesthetized with a mixture of ketamine (90 mg/kg; Hofspira) and xylazine (10 mg/kg; Sigma-Aldrich) via intraperitoneal injection and anticoagulated with heparin (Fresenvis Kabi, Lake Zurich, IL) prior to tissue collection. Hearts were excised, cannulated on a Langendorff apparatus, and perfused in a retrograde manner with Krebs-Henseleit buffer (95% O₂, 5% CO₂; pH 7.4) under constant pressure (100 cmH2O) and temperature (37°C) as previously described for 5 minutes to washout blood¹⁶. Buffer consisted of (in mmol/L): NaCl (120), KCl (4.7), KH₂PO₄ (1.2), NaHCO₃ (25), MgSO₄ (1.2), d-Glucose (14), and CaCl₂ (1.75). Following perfusion, hearts were weighed and either sectioned for histology, or equally halved, and snap-frozen in liquid nitrogen and stored at −8°C for molecular studies. Tissue was collected for prepartum endpoints at E17.5, and postpartum endpoints at P12.5. Cross-fostering of offspring was not performed in this study, and dams were kept with their respective litters until P12.5.

Echocardiography

All echocardiography procedures were performed by an investigator blinded to experimental conditions using a preclinical ultrasound imaging system (Vevo 2100; FUJIFILM VisualSonics, Toronto, ON, Canada) with a 40-MHz linear transducer, as previously described¹⁶. Echocardiography was performed on pregnant mice and embryonic pups under anesthesia using isoflurane. Mice were induced with 3–4% isoflurane and maintained at 1–2% thereafter. Postnatal pups underwent conscious transthoracic echocardiography. The M-mode echocardiogram was acquired from the short-axis view of the left ventricle at the level of the midpapillary muscles (200 m/s sweep speed). Semiautomated continuous tracing of the left ventricular walls from this axis view measured, calculated, or extrapolated the following cardiac parameters using 3–5 cardiac cycles: left ventricular posterior wall thickness at end diastole (LVPWd), left ventricular posterior wall thickness at end systole (LVPWs), left ventricular internal diameter at end diastole (LVIDd), left ventricular internal diameter at end systole (LVIDs), percent fractional shortening (FS), percent ejection fraction (EF), heart rate (H), and cardiac output (CO). Parameters were calculated using the LVTool function in Vevo LAB (VisualSonics, Fujifilm). Embryonic and maternal echocardiography was performed at E16.5 and E17.5, and at days P4 and P5 for pups.

Histology and Cardiomyocyte Cross Sectional Area Measurement

After perfusion and gravimetric measurement, a transverse cross-section of the heart was taken, including both the left and right ventricle. Sections were taken from the middle third of the heart. Tissues were fixed in 4% formalin, embedded in paraffin, sectioned (4μM) and mounted on slides. Heart sections were stained using Hematoxylin and Eosin (H&E) to visualize cardiomyocyte integrity (2 sections) or Masson’s Trichrome to determine the presence of fibrosis (4 sections) and delineate cell boundaries to assess cardiomyocyte cross-sectional area as detailed in prior studies^40–43. Sections were then imaged at 20x using Aperio ScanScope CS (Leica Biosystems, Deer Park, IL). Transverse sections were analyzed for fibrosis expressed as a percentage of total area using an Aperio ImageScope macro (Leica Biosystems, Deer Park, IL). Cross-sectional area of cardiomyocytes was determined from Masson’s Trichrome stained tissue, with 25 measurements taken from one region per transverse section across 4 sections, totaling 100 different measurements per heart. Each cell was outlined by a blinded investigator using Concentric by Proscia, allowing the area of the section to be calculated.

RNA Isolation, Extraction, and cDNA Conversion

Hearts were homogenized with TRIzol (1 mL, Ambion) using a bead-mill tissue homogenizer (2 × 30 s cycles, 0°C, 7,200 rpm; Precellys Evolution 24). Lysates were mixed with chloroform (200 μL, Thermo Fisher), incubated (5 min, 25°C), and centrifuged (12,000 g, 15 min, 4°C) for phase separation. RNA collected from the upper phase was precipitated by incubation (10 min, 25°C) with isopropyl alcohol (500 μL) and centrifugation (12,000 g, 8 min, 4°C). RNA pellets were washed with 75% ethanol, centrifuged (12,000 g, 10 min, 4°C), and air-dried under a laminar flow hood (30 min). Following RNA solubilization (100 μL, DEPC-treated water), RNA concentration and purity (A260/A280 range 1.98 to 2.06) were measured via spectrophotometry (1 mL sample, NanoDrop 100, Thermo Fisher). RNA was subsequently converted to cDNA per manufacturer’s instructions (High-Capacity cDNA Reverse Transcription Kit, 4368814, Thermo Fisher). Briefly, the reverse transcription reaction mix was prepared on ice, RNA (2 μg) was added, and the samples were run on a thermocycler (60 min, 37°C; 5 min, 95°C; held, 4°C; Applied Biosystems); cDNA was stored at −20°C until use.

Quantitative RT-PCR

Expression levels of mRNA transcripts were measured using generated cDNA, a PCR (Polymerase Chain Reaction) master mix (TaqMan Fast Advanced Master Mix, Applied Biosystems) and the following validated primers (TaqMan, Applied Biosystems) on a thermocycler (2 min, 50°C; 2 min, 95°C; (1 s, 95°C; 20 s, 60°C) × 40), (QuantStudio 3, Applied Biosystems, Thermo Fisher) in a 96-well plate (USA Scientific, Beltsville MD): Acta1 (Mm00808218_g1), Akt1 (Mm01331626_m1), Esr1 (Mm00433149_m1), Esr2 (Mm00599821_m1), Gapdh (Mm99999915_g1), Kncd3 (Mm01302126_m1), Nos3 (Mm00435217_m1), Myh6 (Mm00440359_m1), Myh7 (Mm00600555_m1), Nppa (Mm01255747_g1), Nppb (Mm01255770_g1), Pgrmc1 (Mm00443985_m1), Pgrmc2 (Mm01283154_m1), Prlr (Mm04336676_m1), and Vegfa (Mm00437304_m1). Expression was determined using the ΔΔCT method and normalized to Gapdh, which did not change in cycle time with iAs treatment.

Heart Homogenization and Protein Isolation

Hearts were homogenized in cell lysis buffer (1 mL; Cell Signaling Technology, Danvers, MA) supplemented with a protease and phosphatase-inhibitor cocktail (Cell Signaling Technology) using a hard tissue lysing kit (Precellys CK28 Lysing Kit, Bertin Instruments) with a bead-mill tissue homogenizer (2 × 30s cycles, 0°C, 7,2000 rpm; Precellys Evolution 24, Bertin Instruments). Supernatant was collected as total crude homogenate; protein concentration was determined via Bradford assay and aliquots of homogenate were stored at −80°C.

Western Blot

Protein homogenates (30 μg) were separated (20 min, 75 V; 100 min, 175 V) on a gradient Bis-Tris SDS-PAGE gel (4–12%, NuPAGE; Thermo Fisher, Carlsbad, CA) and transferred (90 min, 220 mA) to a PVDF membrane (Thermo Fisher). Every gel included two molecular-weight markers for separate regions of interest (High Range Color-Coded Prestained Protein Marker; Cell Signaling Technology; and Novex Prestained Protein Standard, Thermo Fisher) on opposite ends of the gel. Total protein served as the loading control by covalently labeling lysine residues (No-Stain Protein Labeling Reagent, Thermo Fisher) following the manufacturer’s protocol and visualized by fluorescence imaging (488 nm). Membranes were blocked (1h) with bovine serum albumin (5% wt/vol, Sigma-Aldrich) in Tris-buffered saline with Tween-20 (0.1%), and subsequently incubated (overnight, 4°C) with primary antibodies for p-Akt (S473) (1:1000, Rabbit mAb, Cell Signaling, 4060S), Akt (pan) (1:1000, Rabbit mAb, Cell Signaling, 4691S), p44/42 MAPK (ERK1/2) (1:1000, Rabbit mAb, Cell Signaling, 9102S), and p-p44/42 MAPK (T202/Y204)(1:1000, Rabbit Ab, Cell Signaling, 9101S). Membranes were then incubated with secondary antibodies (Anti-rabbit IgG, HRP-linked Ab, Thermo Fisher) and visualized using chemiluminescence substrate (SuperSignal West Pico PLUS, Thermo Fisher) and an iBright imager (Invitrogen, Thermo Fisher). Membranes were stripped with Reblot Plus Mild Solution (Millipore) when necessary to re-probe total protein following examination of phosphorylation status. No-Stain total protein and full western blot scans are included in Supplemental Figure 2.

Statistical Analysis

Sample sizes of mice for each experiment were estimated a priori via power analysis (power = 0.80, effect size = 0.25, alpha = 0.05) based on data generated in previous studies from our group^15,16. Mice were randomized during collection to minimize batch effects. All offspring data were averaged by litter to minimize litter effect. All data are expressed as the mean ± SEM and were analyzed using GraphPad Prism (La Jolla, CA). Statistical outliers were identified using the ROUT method (Q=1%). Statistical comparisons between groups were determined using an ordinary one-way ANOVA with Sidak’s multiple comparisons test, a two-way ANOVA with Tukey’s multiple comparisons test or a two-tailed Mann-Whitney test as appropriate. Significance was set at p < 0.05.

RESULTS

iAs exposure alters cardiac function in late pregnancy

To determine whether iAs exposure alters maternal cardiac structure or function during pregnancy, transthoracic echocardiography was performed at E16.5 and E17.5 during the late pregnancy period (Fig. 2). We found that pregnant dams exposed to 1000μg/L iAs had an increase in ejection fraction trending towards significance compared to non-exposed pregnant control and 100 μg/L iAs exposed pregnant dams, but no changes were observed between non-pregnant exposed groups (Fig. 2A).Pregnant dams exposed to 1000 μg/L iAs also exhibited a significant increase in fractional shortening compared to control and 100 μg/L iAs exposed dams with no change between non-pregnant exposure groups (Fig. 2B). Additionally, pregnant mice exposed to 1000 μg/L iAs showed a significant increase in cardiac output compared to control and 100 μg/L exposed pregnant dams, as well as 100 and 1000 μg/L iAs exposed non-pregnant females (Fig. 2C). Since cardiac output is the product of heart rate and stroke volume, we examined these two parameters. We found no change in heart rate across pregnant or non-pregnant females (Fig. 2D), but stroke volume was significantly increased in 1000 μg/L iAs exposed pregnant dams compared to non-pregnant females exposed to 1000 μg/L iAs (Fig. 2E). We also found no change in end systolic or diastolic volumes, or in average wall thickness across all pregnant and non-pregnant exposure groups (Figs. 2F–H).

Figure 2: iAs exposure increases maternal cardiac function during pregnancy.

Cardiac parameters as measured by transthoracic echocardiography in both pregnant and non-pregnant mice including: cardiac output (A), stroke volume (B), heart rate (C), ejection fraction (D), fractional shortening (E), end diastolic volume (EDV; F), end systolic volume (ESV; G), average wall thickness (H), left ventricular internal diameter in diastole (LVID;d; I), and left ventricular internal diameter in systole (LVID;s; J) (*= P < 0.05, ** = P< 0.01 vs. control; n=7–9 mice/group). Significance was determined by two-way ANOVA with Tukey’s multiple comparisons test.

iAs exposure alters prepartum hypertrophic cardiac remodeling

Reversible cardiac hypertrophy is known to occur during pregnancy^24,25, so we investigated heart size across all groups to potentially account for our observed changes in cardiac output and ejection fraction. As expected, there was a significant increase in heart size in control pregnant mice compared to control non-pregnant females (Figs. 3A, 3B). Surprisingly, during late pregnancy, we found that maternal hearts were significantly smaller in dams exposed to 1000 μg/L iAs compared to non-exposed pregnant controls and were not significantly different from non-pregnant females exposed to 1000 μg/L iAs (Fig. 3A). Exposure to 100 μg/L iAs, on the other hand, was not sufficient to reduce heart growth relative to non-exposed pregnant controls as heart size was increased as expected compared to non-pregnant mice with the same exposure (Fig. 3B). Next, we examined cardiomyocyte cross-sectional area in histological sections, and consistent with our observed differences in heart size, we found that cardiomyocyte cross-sectional area was significantly less in 1000 μg/L iAs-exposed hearts compared to controls (Figs. 3E, 3F). Masson’s trichrome staining was also performed on transverse heart sections to examine changes in collagen deposition, but no differences were seen in collagen deposition or fibrosis (Figs. 3C, 3D).

Figure 3: iAs exposure blunts maternal cardiac hypertrophy.

Heart weight divided by tibia length in pregnant and non-pregnant females with 1000 μg/L (A) and 100 μg/L (B) iAs exposure compared to non-exposed controls (n=11–18 mice/group; *= P < 0.05, ** = P< 0.01 vs. control). Quantification of fibrosis (C) and representative images (D, scale bar 2mm) in pregnant mice exposed to 1000 μg/L iAs vs. Non-exposed pregnant controls (Ctrl). Quantification of cardiomyocyte cross-sectional area (E) and representative images of cardiomyocyte cross-section (F, scale bar 100 μm) taken from 4 sections per heart with 25 cardiomyocytes per section (n=5 hearts/group, 100 cardiomyocytes/heart; **= P<0.01 vs. control). Significance was determined by two-way ANOVA with Sidak’s multiple comparisons test or Mann-Whitney test.

iAs exposure during pregnancy alters the transcriptional profile of hormone receptors

Hormone signaling is important for cardiovascular adaptation and heart growth during pregnancy, so we examined the expression of various hormone receptors in the maternal heart during late pregnancy. Estrogen impacts hypertrophic remodeling in the pregnant heart^29,44 and we found that exposure to iAs induced a dose-dependent decrease in Esr1 (Estrogen Receptor α, ERα) mRNA levels in the whole heart (Fig. 4A). Esr2 (Estrogen receptor β) was also probed, but the cycle time (CT) values were greater than 38, indicating low levels of expression in whole myocardium. Progesterone is also known to regulate pregnancy-induced hypertrophy^26–28, and we found that exposure to both 100 and 1000 μg/L iAs during gestation downregulated cardiac transcription of the progesterone membrane receptors Pgrmc1 and Pgrmc2 (Figs. 4B, 4C). Cytosolic progesterone receptors were also probed, but levels were too low in whole cardiac tissue to be detected. We also examined Prlr (prolactin receptor) expression in the heart, as prolactin is another hormone that is important during pregnancy, but we found that the levels were not significantly altered (Fig. 4D).

Figure 4: iAs downregulates estrogen receptor alpha transcripts during late pregnancy.

Myocardial mRNA transcript levels of estrogen receptor alpha (ESR1; A), progesterone receptor membrane component 1 (Pgrmc1; B), progesterone receptor membrane component 2 (Pgrmc2; C), and prolactin receptor (Prlr; D) in pregnant females at E17.5 as measured via RT-qPCR (**= P<0.01 vs. control). Transcript levels were normalized to Gapdh mRNA expression, which did not change with iAs exposure. Outliers were determined by the ROUT method (Q=1%), and significance was determined by one-way ANOVA with Sidak’s multiple comparisons test or Mann-Whitney test (n=7–9 mice/group).

iAs exposure and transcriptional markers of hypertrophy

Next, we examined the expression of gene transcripts typically associated with physiologic and pathologic hypertrophy, as well as genes which have been implicated in pregnancy-induced hypertrophy⁴⁵. In hearts exposed to both 100 and 1000 μg/L iAs during pregnancy, we found a decrease in Akt1 (Akt), a major regulator of physiologic hypertrophy⁴⁶, which was trending towards significance (Fig. 5A). We also found a reduction in Nppa (ANP) at both doses (Fig. 5B). Interestingly, we found a significant reduction in Vegfa transcripts at 100 μg/L iAs only (Fig. 5C). Other markers of physiologic and pathologic hypertrophy were not altered (Fig. 5D–I). Transcriptional results were confirmed via western blot and protein levels of Akt were significantly reduced in 1000 μg/L iAs exposed hearts, while Akt phosphorylation (Ser473) relative to total Akt showed a trending decrease (Fig. 6A and 6B). Extracellular-regulated kinase 1 and 2 (ERK1/2) is also a major regulator of pregnancy-induced hypertrophy⁴⁶. However, we found that levels of ERK1/2, as well as phosphorylation of the activation site, were unchanged with iAs exposure (Fig. 6C and 6D).

Figure 5: iAs exposure reduces Akt and ANP transcripts in the prepartum heart.

Myocardial mRNA transcript levels of protein kinase B (Akt; A), natriuretic peptide A (ANP; B), vascular endothelial growth factor (Vegfa; C), potassium voltage-gated channel 4.3 (Kv4.3; D), endothelial nitric oxide synthase (Nos3; E), natriuretic peptide B (Nppb; F), β-myosin heavy chain (Myh7; G), α-myosin heavy chain (Myh6; H), skeletal muscle α-actin (Acta1; I), (*=P<0.05, n=7–9 hearts/group). Outliers were determined by the ROUT method (Q=1%), and significance was determined by a one-way ANOVA with Sidak’s multiple comparisons test.

Figure 6: iAs exposure reduces protein levels of Akt in the prepartum heart.

Myocardial protein expression of phosphorylated Akt at the activating site, S437, normalized to total Akt (A) in dam hearts (P= 0.0776 vs. control, n= 5 hearts/group). Protein expression of total Akt (B) in dam hearts (***P=0.0005 vs. control). Protein expression of phosphorylated ERK1/2 at the activating sites, T202/Y204 (C), and of total ERK1/2 (D). Protein levels of each target were normalized to total transferred protein levels. Outliers were determined by the ROUT method (Q=1%), and significance was determined by a Mann-Whitney test.

In utero exposure to iAs and offspring outcomes

Finally, we examined the impact of gestational iAs on offspring outcomes with a focus on cardiac development. We found that in utero exposure to 100 or 1000 μg/L iAs during gestation resulted in no change to litter size (Fig. 7A) or embryonic body weight (Fig. 7B). However, heart weight relative to body weight was increased at E17.5 after exposure to 1000 μg/L iAs (Fig. 7C). To further delineate changes in cardiac structure and function, echocardiography was performed in utero at E17.5, but we found no change in average wall thickness (Fig. 7D), LVPWd (Fig. 7E), IVSd (Fig. 7F), EF (Fig. 7H), or FS (Fig. 7I). However, we did find an increase in the left ventricular internal diameter during diastole at E17.5 with exposure to 1000 μg/L IAs (LVIDd, Fig. 7G).

Figure 7: Offspring from iAs exposed dams show increased heart weight and left ventricular internal diameter at E17.5.

Litter size (A), body weight (B), and embryonic heart weight normalized to body weight (C), in offspring exposed to 100 and 1000 μg/L iAs during gestation (n= 16–18 litters/group). Cardiac parameters measured in utero using echocardiography including average wall thickness (D), left ventricular posterior wall thickness (LVPW; E), interventricular septum thickness (IVS; F), left ventricular internal diameter (LVID; G), ejection fraction (EF; H), and fractional shortening (FS; I). Outliers were determined by the ROUT method (Q=1%), and significance was determined by one-way ANOVA with Sidak’s multiple comparisons test.

These same parameters were measured again postnatally at P12, and we found that exposure to 100 μg/L iAs in utero reduced body weight at P12 (Fig. 8A). When stratified by sex, both sexes showed a similar significant reduction body weight (Fig. 8B). We also saw no difference in distribution of sex by exposure (Supplemental Table 1). However, while an increase in heart weight was noted at E17.5, we found no change compared to controls at P12, with or without sex stratification (Figs. 8C, 8D).Echocardiography also revealed no change in average wall thickness (Fig. 8E), IVS (Fig. 8F), LVPW (Fig. 8G), LVIDd (Fig. 8H), EF (Fig. 8I), or FS (Fig. 8J).

Figure 8: Offspring exposed to iAs in utero have reduced body weight at P12.

Body weight at P12 (A), body weight stratified by sex (B), heart weight relative to body weight (C), and heart weight relative to body weight stratified by sex (D) measured gravimetrically. Cardiac functional parameters measured via echocardiography including average wall thickness (E), interventricular septum thickness (IVS; F), left ventricular posterior wall thickness (LVPW; G), left ventricular internal diameter (LVID; H), ejection fraction (EF; I), and fractional shortening (FS; J). Significance was determined by one-way ANOVA with Sidak’s multiple comparisons test.

DISCUSSION

We report here for the first time that environmentally relevant iAs exposure during gestation blunts pregnancy-induced growth of the maternal heart during late pregnancy and downregulates hormone receptors ERα and PGRMC1/2, as well as ANP and Akt expression. Pregnancy-induced cardiac hypertrophy is critical for meeting the circulatory demands of the mother and the developing fetus during gestation, and we find that iAs exposure disrupts this important hypertrophic process^17,18. We observed a compensatory change in cardiac function in iAs-exposed dams (Fig. 2), despite the lack of proper heart enlargement (Fig. 3), along with downregulation of ERα, PGRMC1/2, (Fig. 4), ANP and Akt (Figs. 5 and 6). Additionally, we found an increase in heart weight relative to body weight in offspring (Fig. 7) and decreased growth at P12 (Fig. 8) suggesting that either direct iAs exposure or improper maternal cardiac remodeling may contribute to adverse outcomes in offspring.

While our previous study demonstrated that iAs induced sex-dependent changes in cardiac hypertrophy, these changes were exclusive to adult males and adult females remained unaffected after an 8-week exposure¹⁶. However, in the current study, we find that pregnant females are particularly vulnerable to gestational iAs exposure. This study is the first to examine the impact of iAs exposure on maternal cardiac remodeling during pregnancy and our findings identify a critical disruption of pregnancy-induced maternal cardiac growth.

iAs exposure and maternal cardiac function

Heart growth or hypertrophy is a critical process that occurs during pregnancy. Prior work demonstrates that heart weight significantly increases in late pregnancy and post-partum; a non-significant increase in cardiomyocyte cross-sectional area during late pregnancy was also noted in this study, and this increase in cardiomyocyte cross-sectional area became significant postpartum⁴⁷. Consistent with this prior work, our findings demonstrated a significant increase in heart weight in control and 100 μg/L iAs-exposed dams compared to iAs-exposed and non-exposed non-pregnant females (Fig. 3). However, cardiac growth was blunted in maternal hearts exposed to 1000 μg/L iAs during pregnancy, as we noted significantly smaller heart weight with 1000 μg/L iAs exposure compared to control and 100 μg/L iAs exposed dams (Fig. 3). Cross-sectional area was also significantly lower in cardiomyocytes from 1000 μg/L iAs exposed dams compared to non-exposed pregnant controls (Fig. 3), providing further confirmation that pregnant dams exposed to 1000 μg/L iAs during gestation have smaller hearts during late pregnancy. Left ventricular hypertrophy during pregnancy results from volume overload, and a lack of hypertrophy vs. pregnant controls may indicate a reduction in blood volume expansion^17,48. As such, plasma volume will be examined with iAs exposure in future studies at different timepoints during pregnancy.

Furthermore, we found that cardiac function (i.e., fractional shortening) was significantly enhanced in pregnant dams exposed to 1000 μg/L iAs compared to non-exposed pregnant controls or dams exposed to 100 μg/L iAs (Fig. 2), likely as a compensatory mechanism to increase blood flow to the mother and developing fetus. However, it is important to note that we did not observe the expected increase in cardiac output in non-exposed pregnant dams vs. non-pregnant control mice that has been observed previously⁴⁷. Nonetheless, the increase in cardiac function that we did observe with 1000 μg/L iAs-exposed dams may result from several mechanisms that include enhanced cardiomyocyte contractility or increased β-adrenergic stimulation, and future studies will examine these potential compensatory mechanisms. Importantly, sustained alterations to either of these pathways could result in detrimental changes to the heart, which could include a transition from pregnancy-induced hypertrophy to pathological hypertrophy and/or an increased risk for prolonged cardiovascular disease postpartum⁴⁹. As such, postpartum remodeling and specifically the impact of iAs exposure during pregnancy on the postpartum heart, is a critical area that will be addressed in future mechanistic studies. Importantly, we find no alterations in cardiac size or function with iAs exposure alone in non-pregnant females at E17.5, or 15 days of exposure (Figs. 2 and 3), which is consistent with our prior studies where we found no change to cardiac structure and function in adult female mice with iAs exposure at 4 and 8 weeks^15,16. Taken together, our findings suggest that the unique combination of iAs exposure and pregnancy are necessary for increasing cardiac function (Fig. 2) and blunting pregnancy-induced cardiac growth (Fig. 3).

iAs exposure decreases myocardial hormone receptor expression

Pregnancy-induced hypertrophy is governed largely by sex-hormones, such as progesterone and estradiol, and previous research has shown that iAs can act as an endocrine disruptor^50–53. In the current study we found that exposure to 1000 μg/L iAs downregulated cardiac Esr1, Pgrmc1, and Pgrmc2 during late pregnancy (Fig. 4). Prior work has shown that iAs exposure downregulates various steroid hormone receptors and decreases ERα expression in the uterus, leading to alterations in uterine hypertrophy^50,51. Estradiol has been shown to be anti-hypertrophic in mice at high levels and pro-hypertrophic at low levels in cultured cells, mediating the activation of certain downstream signaling pathways that are crucial for cardiac hypertrophy^25,45,46. Exposure to iAs has also been shown to impact progesterone signaling, decreasing serum progesterone in rats, as well as progesterone receptors in culture^54–56. Consistent with these findings, our data showed decreased transcripts of receptors for progesterone (Pgrmc1, Pgrmc2) and estradiol (Esr1) in the heart (Figs. 4A–4C). Interestingly, we observed that doses of 100 and 1000 μg/L iAs downregulated Pgrmc1 and Pgrmc2 at equal levels, while a dose-dependent reduction in Esr1 was noted with 100 and 1000 μg/L iAs.

Dampened estradiol signaling during pregnancy can impact blood volume expansion and the ability of the heart to grow, as pregnancy-induced cardiac hypertrophy is reliant upon major signaling pathways activated by a combination of mechanotransduction due to increased blood volume and hormone signaling^25,46,57. The increase in plasma volume during pregnancy is also a consequence of sodium resorption and water retention, signaled via placental estrogen production⁵⁸. As such, altered Esr1 and Pgrmc1/2 expression in the maternal heart during pregnancy may contribute to our observed changes in cardiac function and heart size in pregnant mice exposed to 1000 μg/L iAs, but further studies are needed to further characterize the impact of iAs on hormonal changes at different time points throughout pregnancy and establish causality with a specific focus on hormone levels.

Akt and ANP reduction in maternal hearts exposed to iAs

Transcriptionally, pregnancy-induced hypertrophy resembles exercise-induced hypertrophy, but there are notable differences²⁴. For example, Acta1, ANP, and BNP are found to be upregulated in late pregnancy^24,59,60. When probing for genes associated with pregnancy-induced cardiac hypertrophy^17,45, we found that dams exposed to either 100 or 1000 μg/L iAs showed reduced transcript levels of Akt1, a primary mediator of pregnancy-induced hypertrophy⁴⁶, in the maternal heart but this change was not significant (Fig. 5A). However, we did find a significant reduction in the protein expression of Akt in the maternal heart, and phosphorylation of the active site showed a trending decrease with iAs exposure (Figs. 6A, 6B). Our findings are consistent with a prior study, which demonstrated that exposure to iAs in adult rats reduced Akt expression in cardiac tissue irrespective of pregnancy⁶¹. While iAs exposure reduced myocardial protein expression of Akt in our model, additional research is needed to determine if this reduction in Akt expression specifically contributes to changes heart function and the blunted hypertrophy observed in 1000 μg/L iAs-exposed pregnant mice.

Exposure to 100 or 1000 μg/L iAs induced a ~50% reduction in transcript levels of Nppa (ANP) in the maternal heart (Fig. 5B). ANP is secreted by the heart in response to atrial distension and is found to be increased consistently in blood plasma throughout pregnancy^59,62,63. ANP is a vasodilator and transcription is induced by atrial stretching due to volume overload⁶⁴. These findings further suggest that iAs exposure could alter plasma volume expansion during pregnancy. In humans, exposure to iAs increased hemoglobin concentration and decreased blood protein concentration, indicating that iAs exposure may affect plasma volume regulation^65,66. However, this has not been explored in the context of pregnancy. While a reduction in both ANP and Akt are associated with cardiovascular alterations during pregnancy, the reduction at both doses and the resulting pathophysiologic changes only at the highest dose may have several implications. While it is possible that ANP and Akt are acting in a synergistic manner with our observed changes in hormone receptor signaling, it is also possible that altered ANP and Akt levels are not associated with the changes in hypertrophic signaling and heart size at this timepoint. However, the full molecular pathway has yet to be elucidated, and thus it will be important to determine how the molecular profile of the maternal heart changes with iAs exposure during early- and mid-pregnancy to further our understanding of iAs exposure on the maternal heart.

Effects of gestational iAs exposure on offspring

In offspring, we found that heart weight was increased with 1000μ/L iAs exposure, but not with 100 μg/L iAs (Fig. 7C). Left ventricular diameter was also increased at E17.5 with 1000 μg/L iAs exposure (Fig. 7G). In mammals, heart cells proliferate from conception until late gestation and then switch to cell hypertrophy shortly after birth⁶⁷. Pregnancies with metabolic dysfunction, such as obesity and diabetes, result in increased fetal cardiac hypertrophy due to increased cardiomyocyte cell size^68–70. In adult hearts, enlarged left ventricular diameter in combination with decreased ejection fraction culminates in heart failure⁷¹. However, we do not see changes to ejection fraction with either iAs exposure dose (Fig. 7H), which suggests that there are only structural changes to the heart at E17.5. Several epidemiologic studies have associated in utero exposure to iAs with congenital heart defect^72–74, but this has yet to be replicated in a murine model with a relevant exposure dose or route of exposure. We also examined postnatal cardiac function at P12 but found no alterations to cardiac function or structure across doses (Fig. 8). These findings suggest that changes observed in the fetal heart at E17.5 may potentially prime the heart for disease as an adult and many epidemiologic studies have found that in utero exposure to iAs induces cardiovascular disease later in life^75–78.

Furthermore, we did not see alterations to litter size, embryonic body weight, or differences in sex distribution of litters at E17.5 with exposure to either 100 or 1000 μg/L iAs beginning at E2.5 (Fig. 7A, 7C). In humans, in utero exposure to iAs has been shown to increase spontaneous abortion and decrease birthweight^79–81. While it is still unclear if this phenotype is seen in mice, it is possible that disrupted embryonic growth is due to exposure occurring before E2.5. We characterized offspring body and organ growth postnatally at P12 and found that exposure to 100 μg/L iAs in utero impaired growth as measured by body weight at P12 (Fig. 8A). Similar findings were observed in a prior study using doses as low as 10 μg/L iAs⁸². However, that study showed that cross-fostering the pups reduced the growth deficit, suggesting that the reduction in growth was due to breast milk nutrition rather than altered in utero programming⁸². Since our model halted iAs exposure at birth, and we did not cross-foster, it is possible that iAs may have lasting effects on breast milk nutrition. Additionally, iAs has been shown to alter epigenetic re-programming in offspring after in utero exposure^83,84. In human studies, iAs has been shown to regulate birthweight via placental-derived miRNAs^85,86. While we see evidence of some phenotype in offspring exposed to iAs in utero, more experiments are needed to define exposure windows and understand how these changes persist into adulthood.

LIMITATIONS

While this study provides important insights into the effects of iAs on maternal health, there are limitations that require acknowledgement. Firstly, the findings of this study are largely descriptive and although we do report molecular changes, additional work is required to establish mechanism and causality. Furthermore, this study used only a single prepartum timepoint, so future studies will be needed to characterize the physiologic growth of the maternal heart during early and mid-pregnancy timepoints throughout the 21-day window of murine pregnancy. We also cannot rule out arsenic-induced atrophy since atrophic signaling was not examined. The exposure was also started at E2.5 during gestation, while a real-world exposure would encompass a lifetime. Even though this is a limitation of our study, this exposure scenario is important for defining a critical window of maternal exposure. Mice were also housed in pairs to reduce stress during pregnancy, and water and food consumption was measured per cage, so individual variability is not accounted for and is a limitation. Additionally, experiments were conducted using a mouse model, and while mice are well characterized and commonly utilized for studies examining pregnancy and iAs exposure, it is important to note that iAs metabolism in mice is different compared to humans^87–89. The mouse metabolizes iAs more quickly than humans, making higher doses in mice equivalent to lower doses in humans. There are also key differences in murine pregnancy compared to that of humans. Many rodent studies have not been able to replicate some of the findings associated with in utero exposure of iAs on development in humans, indicating that there are differences between the two species, and this should be taken into consideration when interpreting these findings. iAs has also been shown to target mitochondria and affect cellular metabolism^90–95. Shifts in cardiac metabolism during pregnancy have been shown to be essential for cardiac hypertrophy^47,96–100. While we did not examine iAs-dependent effects on myocardial metabolism in the current study, these pathways need to be further investigated. Additionally, we found that cardiac function was dampened after the administration of isoflurane, which is consistent with previous studies^101,102. Since isoflurane reduced baseline cardiac function, it is important to note that isoflurane could affect our experimental findings, but all groups were exposed to the same amount of isoflurane during echocardiography. Indeed, the lack of increased cardiac output in non-exposed dams vs. non-pregnant control mice may very well result from the masking of changes in cardiac function by isoflurane, and as such is a limitation of this study that will be further investigated in future studies. Our use of M-mode echocardiography (vs. B-mode), may also affect our ability to resolve structural and functional differences in the heart with pregnancy. Finally, we only examined two iAs doses in our study, 100 μg/L and 1000 μg/L iAs. Lower doses should be used in future studies to understand if these effects, specifically postpartum, persist at lower doses, such as the recommended limit of 10 μg/L provided by the Environmental Protection Agency.

CONCLUSION

iAs exposure during gestation blunted pregnancy-cardiac growth and increased cardiac function during late pregnancy, and led to the downregulation of ERα, PRMGC1/2, ANP and Akt. As such, these changes have the potential to contribute to adverse pregnancy-related cardiovascular events, impact developing offspring, as noted, and may lead to increased risk for maternal CVD postpartum. Taken together, our findings suggest that exposure to iAs during pregnancy causes adverse effects on both the pregnant dam and the offspring. Our findings also add to the growing body of evidence implicating iAs as a driver of cardiovascular disease and a potential risk factor for maternal and fetal health during pregnancy, underscoring the importance of removing iAs from human consumption.

HIGHLIGHTS.

• Arsenic exposure blunts maternal cardiac hypertrophy during late pregnancy

• Arsenic exposure alters myocardial hormone receptor and hypertrophic gene expression in late pregnancy

• Offspring from this exposure model showed enlarged hearts at embryonic day 17.5 and reduced body weight at P12