Maternal nicotine exposure induces congenital heart defects in the offspring of mice

Abstract Maternal cigarette smoking is a risk factor for congenital heart defects (CHDs). Nicotine replacement therapies are often offered to pregnant women following failed attempts of smoking cessation. However, the impact of nicotine on embryonic heart development is not well understood. In the present study, the effects of maternal nicotine exposure (MNE) during pregnancy on foetal heart morphogenesis were studied. Adult female mice were treated with nicotine using subcutaneous osmotic pumps at 0.75 or 1.5 mg/kg/day and subsequently bred with male mice. Our results show that MNE dose‐dependently increased CHDs in foetal mice. CHDs included atrial and ventricular septal defects, double outlet right ventricle, unguarded tricuspid orifice, hypoplastic left ventricle, thickened aortic and pulmonary valves, and ventricular hypertrophy. MNE also significantly reduced coronary artery size and vessel abundance in foetal hearts. Moreover, MNE resulted in higher levels of oxidative stress and altered the expression of key cardiogenic regulators in the developing heart. Nicotine exposure reduced epicardial‐to‐mesenchymal transition in foetal hearts. In conclusion, MNE induces CHDs and coronary artery malformation in mice. These findings provide insight into the adverse outcomes of foetuses by MNE during pregnancy.

To reduce the prevalence of maternal smoking, pharmacotherapies such as nicotine replacement therapies (NRTs) are often offered to pregnant women following failed attempts at smoking cessation. 9,10 Another popular method for smoking cessation is nicotine containing electronic cigarettes (e-cigarettes). In young adults of reproductive age (20)(21)(22)(23)(24)(25)(26)(27)(28), the use of e-cigarettes is rapidly increasing. 11,12 Although NRTs and e-cigarettes contain nicotine, they are perceived to be less harmful than cigarette smoking since they do not contain additional toxins found in tobacco smoke. The relationship between NRTs/e-cigarettes and congenital anomalies is poorly understood. Additional studies are required to elucidate the nicotine's effects on foetal development. 13 Currently available clinical evidence on the foetal safety of maternal use of NRTs during pregnancy is discordant. [14][15][16][17] It remains unclear whether these replacements are safe for the foetuses.
Maternal nicotine exposure (MNE) during pregnancies is more likely to have adverse outcomes. 18 In animal studies, MNE adversely affects various organs during foetal development, including exacerbation of fibrosis of lungs and kidney, alterations in brain structure and function, mitochondrial dysfunction in the pancreas and impairment of structure and function of the placenta. 19 In foetal hearts, nicotine delays the development of the sinoatrial node and induces cardiac arrhythmia. 20,21 Further, nicotine reduces blood oxygenation and increases blood pressure in foetuses. 18 Increases in oxidative stress may contribute to these detrimental effects of foetal nicotine exposure. 22 Reactive oxygen species (ROS) such as superoxide and hydrogen peroxide mediate many fundamental cellular processes critical to embryonic development including cell differentiation, proliferation, migration and programmed cell death. 23,24 Elevated ROS levels induce oxidative stress, which alters genetic expression and disrupts developmental processes. In foetal hearts, oxidative stress contributes to the pathogenesis of CHDs and coronary artery malformation. 25,26 To our knowledge, there is no report on the direct effects of MNE on the pathogenesis of CHDs. We hypothesized that MNE during pregnancy would impair foetal heart development, resulting in CHDs and hypoplastic coronary arteries. We tested this hypothesis using subcutaneous osmotic pumps administering clinically relevant doses of nicotine to female mice during gestation.

| Animals
This study used mice in accordance with the Guide to Care and Use of Animals of the Canadian Council of Animal Care and was approved by the Animal Care Committee at Western University, Canada.
C57Bl/6 mice were purchased from Charles River Laboratories, Canada. All efforts were made to minimize the number of animals used and to minimize their suffering. Animals were housed in a 12 h light/dark cycle and had access to standard chow and water. Females at 8-10 weeks of age had osmotic pumps (Alzet #2004,) implanted subcutaneously, releasing nicotine at 0.75 or 1.5 mg/kg/day. These doses mimic a light smoker, of 1-10 cigarettes per day. [27][28][29] Control mice did not receive pump implantation. Fourteen days after implantation of the pump, the females were bred overnight in cages with healthy males, then returned to their original cage in the morning. The presence of a vaginal plug indicated embryonic day (E) 0.5.
Females that had an unsuccessful pregnancy following the presence of the plug were bred again with males. After 2 unsuccessful pregnancies, females were sacrificed and not included in analysis.
Embryos from pregnant mice with or without maternal nicotine exposure (MNE) were collected at E10.5 to E18.5 via caesarean section under ketamine (50 mg/kg, IP) and xylazine (10 mg/kg, IP) anaesthesia. Dams were sacrificed by cervical dislocation after embryos were collected. Embryonic/foetal hearts were harvested at E10.5 to analyse mRNA and oxidative stress during embryonic development; at E12.5 to analyse epicardial EMT potential, and at E18.5 to analyse morphology and mRNA. For mRNA analysis, hearts were flash frozen in liquid nitrogen and stored at −80°C freezer.

| Osmotic mini pumps
Alzet (#2004) 28-day osmotic minipumps were prepared 24 h prior to implantation. The online Alzet pump calculator was used to determine the dose with which to fill the pump with nicotine ((-)-nicotine N3876, Sigma-Aldrich) solution according to the weight of the animal. An intramuscular injection of a mixture of ketamine (25 mg/ml), xylazine (2.5 mg/ml) and atropine (30 µg/ml) was used to anaesthetize the mice for surgical pump implantation. A small incision was made slightly posterior to the scapulae, and the pump was implanted.
The incision was closed with silk sutures, with the mice monitored throughout recovery. To relieve pain, buprenorphine (0.05 mg/kg, s.c., q8h) was used for 2 days postsurgery.

| Histological Analysis of CHD
Neonates (P0) and foetuses at E17.5 and E18.5 were collected for histological analysis. To collect foetuses, pregnant mice were anaesthetized with an intramuscular injection anaesthetic mixture (as above), and a caesarean section was performed. 30 The upper torso of foetal and neonatal mice were isolated and immediately fixed in 4% paraformaldehyde for 18 h at 4 °C, then dehydrated in ethanol and embedded in paraffin. Samples were serially sectioned at 5 μm thickness using a Leica RM2255 microtome. Transverse sections started at the level of the thymus (just above the aortic arch) and continued until after the apex of the heart. Heart sections were stained with haematoxylin/eosin (H/E) to diagnose CHDs during blinded examination under a light microscope (Zeiss Observer D1,).
All quantifications of histological images were performed using ZEN microscope software (Zeiss). The diameters of the proximal left and right coronary arteries were measured at 50 μm from the left coronary artery ostium and on a section 250 μm below the left coronary artery ostium where the right coronary artery was the most prominent, respectively.

| Immunohistochemistry
E18.5 foetal heart sections were immunostained using antiαsmooth muscle actin primary antibody (1:3000 dilution, Abcam,) to visualize coronary arteries. 31 (Table S1). An Eppendorf MasterCycler Realplex (Eppendorf) was used to amplify qPCR mixtures for 35 cycles at temperatures set in accordance with the primer melting temperatures. The Ct values of target genes were normalized to 28S ribosomal RNA using a comparative Ct method.

| Analysis of superoxide, lipid peroxidation and cell proliferation
Frozen E10.5 hearts from both control and MNE groups were sectioned into 8 μm sections using a cryostat (CM1950, Leica).
Dihydroethidine (DHE, 2 μM), a molecular probe for superoxide, was used to measure relative levels of ROS by quantifying fluorescence densitometry in the absence or presence of 100 units/ml of superoxide dismutase (SOD). Fluorescence was imaged using a microscope (Observer D1, Zeiss). Three to 5 images were taken from 5 different sections per heart sample at a fixed exposure time. AxioVision Microsoft software (Observer D1, Zeiss) was used to quantify the signal intensity per area of myocardium.

| Epicardial EMT assay
To determine the effect of nicotine on epicardial epithelial-tomesenchymal transition (EMT), ventricles of E12.5 embryos were harvested, cut into smaller fragments, and plated on the hydrated collagen gel. M199 medium (Sigma-Aldrich) was then added to the culture either with or without nicotine (100 ng/ml, (-)-nicotine N3876 Sigma-Aldrich). After 3 days, images were captured using a phase contrast microscope (Observer D1, Zeiss), and the number of spindle-shaped cells which had grown outward from the explanted ventricles were quantified as previously described. 26,31

| In utero Echocardiography
Directly prior to pup collection, foetal heart function was measured at E18.5 using ultrasound imaging (Vevo 2100) with an MS 750 transducer (VisualSonics). 32 Maternal mice were pre-anaesthetized in a chamber, using 3.0% isoflurane, and subsequently secured in the supine position on a heated dock (temperature 37°C) with their noses in a cone used to deliver 0.5-1.5% isoflurane (for anaesthesia maintenance). An incision was made on the lower abdomen to expose the embryotic sacs. M-mode echocardiography images of the foetal hearts were recorded in the short-axis view. The end-diastolic left ventricular internal diameter and end-systolic left ventricular internal diameter were measured. Ejection fraction and fractional shortening were calculated. Mothers were then anaesthetized with the same mixture as above, and removed from isoflurane for pup collection.

| Statistical analysis
Data are presented as means ± SEM. Statistical analysis was performed between control and MNE using unpaired Student's t-test (GraphPad Prism, version 5.0). A Chi-square or Fisher's exact test was used to analyse the incidence of CHDs and coronary artery malformation. Two-way analysis of variance (ANOVA) followed by Tukey's multiple comparisons test were used for the analysis of sex differences in the CHD incidence and the DHE data. A p value less than 0.05 was considered statistically significant.

| Effects of MNE on foetal weight, placental weight, fertility rate and cardiac function
Foetuses and placenta were weighed at collection. While the weight of placenta was similar between control and MNE (1.5 mg/kg/day) groups (p=n.s), foetal weights and foetal-to-placental weight ratios were significantly lower with MNE (p < 0.05, Figure 1A-C). The rate of successful pregnancy (the per cent of vaginal plugs that lead to a successful pregnancy) was lower in MNE (1.5 mg/kg/day) in comparison with controls (p < 0.01, Figure 1D). The per cent of absorbed pups had a higher trend in the MNE group (8.1%) than the control group (2.9%, p = 0.0548 by Fisher's exact test). Cardiac function of foetuses was assessed by echocardiography in utero at E18.5.
The ejection fraction and fractional shortening of the left ventricle F I G U R E 1 Effects of MNE (1.5 mg/kg/day) on foetal body weight, placental weight, pregnancy rate, cardiac function and CHD sex ratio. Foetal weight (A), placental weight (B) and foetal/placental weight ratio (C) at E17.5 in control (n = 14) and nicotine (n = 29) groups. (D) Fertility rate (number of successful pregnancies to total vaginal plugs). Left ventricular ejection fraction (E) and fractional shortening (F) were determined using echocardiography in utero at E18.5. (G) Percent of male and female foetuses with or without a CHD. An unpaired Student's t-test (A-C, E and F) or chi-squared test (D and G) was performed for statistical analysis. *p < 0.05, **p < 0.01 vs. control (LV) were lower in the MNE group in comparison with control foetuses (p < 0.01, Figure 1E-F), indicating reduced cardiac function in nicotine-exposed foetuses. The sex of foetuses was identified by the Y chromosome specific gene using PCR analysis. Our data showed that while the male to female ratio was about 50% in normal foetuses, CHDs were more prevalent in males compared to females (67% vs. 33%, p < 0.05, Figure 1G).

| MNE induces congenital heart defects in mice
Foetuses from MNE developed CHDs ranging from atrial and ventricular sepal defects (ASD and VSD), double outlet right ventricle (DORV), hypoplastic left ventricle, unguarded tricuspid orifice, trabeculation defect, thickened pulmonary and aortic valves to coronary artery malformation (Table 1). MNE dose-dependently increased the incidence of CHDs in foetuses from 0% in controls to 13% and 33% in 0.75 and 1.5 mg/kg/day nicotine treatments, respectively (p < 0.001). Figure

| MNE induces hypoplastic coronary arteries in mice
Coronary arteries were identified on E18.5 heart sections by α-  (Table 1). Foetuses with both CHD and coronary artery malformation were 4% and 16% in 0.75 and 1.5 mg/ kg/day MNE, respectively.

| MNE changes gene expression in foetal hearts
To study the effects of MNE (1.5 mg/kg/day) on the expression of genes critical to embryonic heart development, foetal hearts were collected at E10.5 and RT-qPCR analysis was performed. Changes in gene expression on pathways responsible for angiogenesis, cell proliferation, differentiation and EMT are summarized in Table 2. MNE hearts in comparison with the controls (p < 0.05), while no significant changes were found in Tbx18, Gata4, Nkx2.5, PKCi or Tbx5 mRNA levels ( Table 2).
To determine signalling pathways on cardiac hypertrophy during later stages of heart development, gene expression was assessed in E18.5 hearts using RT-qPCR analysis. The levels TGF-β1, CyclinD1, β-MHC and BNP mRNAs were significantly higher in MNE hearts (1.5 mg/kg/day) in comparison with the controls (p < 0.05, Table 2).

| MNE lowers cell proliferation
Cell proliferation is critical to heart morphogenesis and coronary artery development. E10.5 hearts were immunostained with the phosphorylated histone H3 (pHH3), a biomarker of proliferating cells ( Figure 4A,B). The pHH3 positive cells were quantified on 4-5 sections per heart. The number of pHH3 positive cells was significantly lower in the ventricular myocardium of the MNE (1.5 mg/kg/day) group in comparison with the control group (p < 0.01, Figure 4C).

| Nicotine exposure inhibits epicardial EMT ex vivo
Epicardial EMT is essential to coronary artery development. To study the effect of nicotine on epicardial EMT, E12.5 heart explants were seeded on collagen coated wells to allow epicardial cell outgrowth and EMT to become spindle-shaped, mesenchymal like cells ( Figure 4D-F). The area of outgrowth and the number of spindleshaped cells were quantified and normalized to heart explant area ( Figure 4G,H). Nicotine-treated heart explants showed significantly less outgrowth and fewer spindled-shaped cells in comparison with the controls (p < 0.05, Figure 4G,H). These data suggest that nicotine exposure inhibits epicardial EMT of the foetal heart.

| MNE increases oxidative stress in foetal hearts
To assess the effect of MNE on oxidative stress in foetal hearts, superoxide levels were determined in E10.5 hearts using dihydroethidine (DHE) as a probe. To verify that the DHE fluorescence was superoxide, some heart sections were incubated with a superoxide scavenger, superoxide dismutase (SOD). Fluorescence images were taken on 5 sections per heart for all groups at the same exposure, and intensity of the fluorescence was quantified using densitometry ( Figure 5A-D). The results showed that hearts with MNE (1.5 mg/ kg/day) had significantly higher superoxide levels than the control group (p < 0.05), which was attenuated by SOD (p < 0.001, Figure 5G).
To assess oxidative damage, levels of 4-hydroxynonenal, a product of lipid peroxidation were determined in E10.5 hearts using immunostaining. Fluorescence images were taken at the same exposure on 5 sections per heart in control and NME (1.5 mg/kg/day) groups ( Figure 5E,F). Our data showed that the MNE hearts had significantly higher levels of lipid peroxidation in comparison to the control group (p < 0.001, Figure 5H).

| DISCUSS ION
Defining safety guidelines of substance use during pregnancy is a critical focus in preventative healthcare to lower the incidence of congenital defects. A risk factor such as MNE could be controlled by establishing updated guidelines for NRT use during pregnancy.
Therefore, we evaluated MNE's effects on heart development and underlying pathways. MNE dose-dependently induced both CHDs and altered coronary artery development in the foetuses of mice.
MNE resulted in higher ROS levels in developing hearts. High ROS levels contribute to CHDs and malformation of coronary arteries in mouse offspring. 36   hearts than those of control hearts. In utero cardiac hypertrophy is due to cardiomyocyte proliferation and hypertrophy. 41 Cyclin D1 has important roles for proliferation and hypertrophic regulation. 42 Increased ventricular β1 are associated with cardiac hypertrophy. 43,44 Upregulation of β-MHC is a marker of cardiac hypertrophy. 45 Our results in mice support the conclusion that MNE causes ventricular hypertrophy. Additionally, haemodynamic expressions of BNP and TGF-changes due to thickened semilunar valves, narrowed orifice of the aorta and pulmonary artery, and septal defects may also contribute to ventricular hypertrophy in MNE hearts.
Excessive production of ROS has been shown to be detrimental to embryogenesis, and oxidative stress promotes the development of CHDs. 26,36 In our study, MNE hearts at E10.5 had higher superoxide and lipid peroxidation levels, indicating oxidative damage. 46 Nicotine binding to nicotinic acetylcholine receptors results in increased ion influx, which increases intracellular Ca 2+ and subsequently disturbs intracellular signalling and organelle function. 47 Increased intracellular Ca 2+ in the foetus causes mitochondrial production of ROS production. 48 Increased mitochondrial ROS production disrupts the endogenous antioxidant capacity which can lower cell proliferation. 36 During embryonic development, high ROS produces DNA damage, and oxidation of proteins and lipids, as well as affecting cellular apoptosis, proliferation, differentiation and inflammation. 36 Increased ROS also reduces nitric oxide bioavailability and eNOS coupling. 30  ml peak plasma levels in humans with a plasma half-life of less than 20 min and a distribution volume 2.6 times body weight, indicating nicotine quickly gets into the tissues after smoking. 28,29 Plasma halflife of nicotine is less than 7 min in mice. 52 A limitation of our study is that we were unsuccessful in measuring plasma nicotine levels in pregnant mice. Thus, we are unable to correlate plasma nicotine levels with CHDs. Also, we did not implant osmotic pumps in control mice, which may interfere with pregnancy. However, female mice were bred 14 days after pump implantation. Thus, any adverse effect of the surgery or pump on pregnancy would be minimal.
In conclusion, in mice, MNE dose-dependently induces CHDs and malformations of the coronary arteries. In embryonic hearts,

ACK N OWLED G EM ENTS
This work was performed in partial fulfilment of the requirements