Fetal e-cigarette exposure programs a neonatal brain hypoxic-ischemic sensitive phenotype via altering DNA methylation patterns and autophagy signaling pathway

Abstract

Maternal e-cigarette (e-cig) exposure is a pressing perinatal health concern. Emerging evidence reveals its potential adverse impacts on brain development in offspring, yet the underlying mechanisms are poorly understood. The present study tested the hypothesis that fetal e-cig exposure induces an aberrant DNA methylation profile in the developing brain, leading to alteration of autophagic flux signaling and programming of a sensitive phenotype to neonatal hypoxic-ischemic encephalopathy (HIE). Pregnant rats were exposed to chronic intermittent e-cig aerosol. Neonates were examined at the age of 9 days old. Maternal e-cig exposure decreased the body weight and brain weight but enhanced the brain-to-body weight ratio in the neonates. E-cig exposure induced a gender-dependent increase in hypoxic-ischemia-induced brain injury in male neonates associated with enhanced reactive oxygen species (ROS) activity. It differentially altered DNA methyltransferase expression and enhanced both global DNA methylation levels and specific CpG methylation at the autophagy-related gene 5 (ATG5) promoter. In addition, maternal e-cig exposure caused downregulations of ATG5, microtubule-associated protein 1 light chain 3β, and sirtuin 1 expression in neonatal brains. Of importance, knockdown of ATG5 in neonatal pups exaggerated neonatal HIE. In conclusion, the present study reveals that maternal e-cig exposure downregulates autophagy-related gene expression via DNA hypermethylation, leading to programming of a hypoxic-ischemic sensitive phenotype in the neonatal brain.

INTRODUCTION

Electronic cigarettes (e-cig), a battery-powered nicotine delivery system (ENDS), became available in the United States market in 2007 (1). Since its introduction into the commercial market, e-cig vaping has obtained wide popularity in general population, especially in pregnant women and adolescents (1, 2). One recent survey conducted between 2007 and 2017 indicated that up to 15% of pregnant women in the United States reported using e-cigarettes (3, 4). Thus, e-cig exposure and abuse in pregnant women have become an increasing health concern in the 21st century. Compared with conventional combustible cigarette smoking, the e-cigarette contains other unique constituents beyond nicotine, such as propylene glycol (PG), vegetable glycerin (VG), formaldehyde, acrolein, flavoring chemicals, heavy metals, and other trace elements, some of which may confer potential toxicity to the developing fetus and offspring (1). Recent studies have shown that e-cig vaping can alter cardiovascular function (5) and human embryonic stem cells (6, 7). Furthermore, it has been shown that e-cig exposure during pregnancy impairs embryo development and inhibits fetal growth (6, 8). These findings suggest that the developing fetus may be uniquely susceptible to e-cigarette-induced toxicity. Nonetheless, little work has been done to determine whether and how this fetal e-cig exposure-mediated detrimental effect may persist into postnatal life and cause a dysfunctional phenotype in offspring.

Hypoxic-ischemic encephalopathy (HIE) remains a leading cause of neonatal morbidity and mortality globally, with incidence rates ranging from 1 to 6 per 1,000 term newborns, and presents with a range of diverse and severe long-lasting neuropsychiatric deficits, such as cerebral palsy, seizure, and cognitive retardation in infants and children (9–11). Neonatal brain ischemic disease etiologies are different from those in adults. It is not only linked to maternal and fetal conditions but also associated with complications of pregnancy. Epidemiological and animal studies suggest that maternal cigarette smoking/nicotine exposure during pregnancy is one of the major risk factors contributing to the incidence of brain ischemic disorder in newborns (10, 12). Our previous studies have demonstrated that perinatal nicotine exposure alters fetal brain development leading to programming of a brain hypoxic-ischemic-sensitive phenotype in postnatal life (13, 14). However, because of the nascent stage of e-cig consumerism, there is limited data to show whether and how e-cigarette exposure during pregnancy causes fetal programming of neonatal ischemic disease.

Autophagy plays a key role in pathophysiological processes. The autophagy pathway is mediated by a group of autophagy-related genes (ATGs). ATG5 is a protein essential for the early stages of autophagosome formation, and recent studies suggest that ATG5 has a crucial role during early embryonic brain development (15). In addition, ATG5 deficiency leads to reactive oxygen species (ROS) accumulation, resulting in mitochondrial dysfunction and brain injury (16). Previous studies have shown that cigarette smoke/nicotine exposure causes tissue and cell damage via the autophagy pathway (17, 18). These studies led us to hypothesize that ATG5/autophagy deficiency is a potential key linker contributing to e-cig exposure-induced brain ischemic-sensitive phenotype in offspring.

In the present study, we used a chronic intermittent e-cigarette aerosol exposure (CIEC) pregnant rat model to examine the effects of CIEC during pregnancy on neonatal growth and brain development. We also used an established HIE model to determine whether fetal e-cig exposure enhances hypoxic-ischemia-induced brain injury in neonates. To determine whether the e-cig-mediated neonatal brain dysfunction is regulated by an epigenetic regulatory mechanism, we examined the global and ATG5 gene-specific DNA methylation patterns in the neonatal brain. Finally, we examined the autophagy signaling-associated gene and protein abundances in the neonatal brain to determine their potential contributions to e-cigarette exposure-induced programming of a brain hypoxic-ischemic sensitive phenotype. Our present study not only presents new evidence for the potential neurotoxicity of maternal e-cigarette exposure on offspring but also confers new insights into the molecular epigenetic mechanisms linking maternal e-cigarette vaporing to the heightened neonatal HIE vulnerability in offspring.

MATERIALS AND METHODS

Animal Model of Chronic Intermittent e-Cigarette Aerosol Exposure

Time-dated pregnant Sprague-Dawley rats were purchased from Charles River Laboratories (Portage, MI). All of the procedures and protocols in the proposed animal studies were approved by the Institutional Animal Care and Use Committee of Loma Linda University and followed the guidelines by the National Institutes of Health Guide for the Care and Use of Laboratory Animals. A chronic intermittent e-cigarette (CIEC) exposure pregnant rat model was established in our laboratory with slight modification from previous studies (5, 19). Briefly, the e-cigarette rodent exposure system was obtained from AutoMate Scientific, Inc. CA. The e-cig exposure system consists of a cylindrical free-moving exposure chamber with e-cig holders, solenoid air valves, and an apparatus of both hardware and software that controls the timing and duration of e-cig activation, as well as schedules multiple deliveries per day for chronic intermittent e-cig inhalation exposure. To mimic the phenomenon of chronic intermittent e-cig exposure in human vapers, we have optimized the system to deliver e-cig aerosol in puff duration of 4 s, 3 puffs in an interpuff interval of 30 s per vaping episode, and one episode per hour in the dark phase of 12 h each day, in which the nicotine and cotinine blood pharmacokinetics of pregnant rats are similar to human e-cigarette users (20, 21).

The time-dated pregnant rats were randomly divided into two groups: the first group was exposed to e-cig (2.4% nicotine) aerosol and the second group was exposed to fresh air in the same exposure chamber (control). We used commercial (BluPlus Cig) e-cigarettes for this project that reflects what real-world e-cig users are experiencing. The dam was exposed to e-cigarettes from gestational day 4 (E4) to gestational day 20 (E20), a total of 17 days, with a vaporing cycle of 12 h per day. The rationale for starting on E4 is that this is the time just before embryo implantation onto the uterine wall in rats. Because rats are nocturnal animals, CIEC inhalation exposure experiments were scheduled in the dark phase of the 12/12 h light/dark circadian cycle. During the light phases of 12 h, rats were returned to their home cages and no e-cig aerosol was delivered. After CIEC exposure, dams were transferred to their home cages for a natural delivery. The neonatal offspring at postnatal day 9 (P9) were used for this project study.

Hypoxic-Ischemic Encephalopathy Model

A modified Rice–Vannuci model was conducted on P9 pups (22). Briefly, pups were anesthetized with isoflurane (4%–5% for induction, 2%–2.5% for maintenance). After sanitization, a neck incision was made slightly off the midline of the anterior neck. The right common carotid artery was isolated from the surrounding structures. Two ligations were implemented on the right common carotid and then the artery was cut. Pups were allowed to recover for 1 h on a heating pad and then transferred to a hypoxic chamber (8% oxygen) for 2 h. Inhalation of 8% oxygen-balanced nitrogen was used to produce systemic hypoxia in the pups while maintaining a survivable severity for later physiological measurement. After hypoxic-ischemic encephalopathy (HIE), pups were returned to their respective dams and monitored closely for any signs of distress, failure to thrive, infection, or serious disability. Buprenorphine was administered for pain relief if necessary.

Intracerebroventricular Injection

P7 pups were administered ATG5 siRNA (E-095125-00-0005, SMARTpool: Accell ATG5 siRNA, Dharmacon) or siRNA scramble (Exiqon) via intracerebroventricular injection, as previously described (13, 23, 24). ATG5 siRNA and siRNA scramble control were prepared according to the manufacturer’s instructions. Pups were anesthetized with isoflurane and placed in a stereotaxic frame (Stoelting Co., Wood Dale, IL). An incision was made on the skull surface and bregma was exposed. A small burr hole (0.5 mm diameter) was drilled into the right skull with the coordinates relative to bregma: 1.5 mm lateral, 2.0 mm posterior. Totally, 100 pmol ATG5 siRNA or siRNA scramble with a total volume of 2 μL were injected at a rate of 1 μL/min with a 10-μL syringe (Stoelting) into the right lateral ventricle following the coordinates relative to bregma: 2.0 mm posterior, 1.5 mm lateral, and 3.0 mm below the skull surface. The injection lasted 2 min and the needle was kept for an additional 5 min and then slowly withdrawn over 5 min to prevent backflow. The burr hole was sealed with bone wax and the wound was closed with sutures.

Infarct Size Measurement

Forty-eight hours after HIE procedures, neonatal pups were euthanized. Coronal slices of the brain (2-mm thick) were cut and immersed in a 2% solution of 2,3,5-triphenyltetrazolium chloride monohydrate (TTC) at 37°C for 5 min and then fixed by 10% formaldehyde overnight for measurement of infarct size. The stained red colors in the brain represent viable areas and the white colors represent infarct areas. Each slice was weighed and photographed separately. The percentage of infarction area for each slice was analyzed by ImageJ software (v. 1.40; National Institutes of Health, Bethesda, MD), corrected by slice weight, summed for each brain, and expressed as a percentage of whole brain.

Reactive Oxygen Species Quantification

Total reactive oxygen species (ROS) production levels in the right hemisphere brain tissue samples were measured with the Oxiselect in vitro reactive oxygen and/or nitrogen species (ROS/RNS) assay kit (Cell Biolabs Inc., San Diego, CA), following the manufacturer’s instructions. As previously described (25–27), brain tissues were homogenized at 50 mg/mL in phosphate-buffered saline (PBS) on the ice and centrifuged at 10,000 rpm for 5 min at 4°C. Fifty microliters of the samples or standard were added to a 96-well plate and mixed with 50 µL of catalyst and 100 µL of 2′,7′-dichlorodihydrofluorescein diacetate (DCF). After incubation at room temperature for 30 min, the fluorescence (Ex480nm/Em530nm) was measured using a Synergy HT Multi-Mode Microplate Reader (Bio-Tek Instruments, Inc., Winooski, VT).

Western Blotting Analysis

Whole right hemisphere brain samples were isolated from P9 rat pups and were homogenized with a lysis buffer containing 150 mmol/L NaCl, 50 mmol/L Tris·HCl, 10 mmol/L EDTA, 0.1% Tween 20, 1% Triton, 0.1% β-mercaptoethanol, 0.1 mmol/L phenylmethylsulfonyl fluoride, 5 µg/mL leupeptin, and 5 µg/mL aprotinin, pH 7.4. Homogenates were centrifuged at 4°C for 15 min at 14,000 g. The supernatants were then collected and protein concentrations were determined via a protein assay kit (Bio-Rad, Hercules, CA). Samples with equal amounts of protein (30 µg) were loaded onto 10% polyacrylamide gel with 0.1% sodium dodecyl sulfate and separated by electrophoresis at 100 V for 90 to 120 min. Proteins were then transferred onto nitrocellulose membranes. Nonspecific binding sites were blocked for 3 to 4 h at room temperature in a Tris-buffered saline solution containing 5% dry milk. The membranes were probed with primary antibodies against ATG5 (1:1,000, Cell Signaling Inc., Cat. No. 12994S), LC3β (1:1,000, Cell Signaling Inc., Cat. No. 2775S), DNMT1 (1:1,000, Cell Signaling Inc., Cat. No. 5032S), DNMT3β (1:1,000, Novus Biologicals Inc., Cat. No. NBP2-80402), NOX2 (1:1,000, Bosterbio Inc., Cat. No. PA1667), and Sirt1 (1:1,000, Cell Signaling Inc., Cat. No. 8469S), respectively, as described previously (13, 14). After washing was completed, membranes were incubated with secondary horseradish peroxidase-conjugated antibodies. Proteins were visualized with enhanced chemiluminescence reagents, and blots were exposed to Hyperfilm. The results were analyzed with Kodak ID image analysis software. Band intensities were normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH).

Global Methylation Level Quantification

The global DNA methylation levels in brain tissues were determined by measuring 5-methylcytosine (5-mC) using a 5-mC DNA ELISA Kit (Zymo Research), following the manufacturer’s instructions. Briefly, as described previously (25, 28), genomic DNA was isolated from the right hemisphere of the brains on P9. A total of 100 ng of each DNA sample and negative/positive controls were denatured and used to coat the plate wells with 5-mC of coating buffer. After incubation at 37°C for 1 h, these wells were washed three times with 5-mC ELISA buffer and then incubated again at 37°C for 30 min. Antibody mix consisting of anti-5-methylcytosine and secondary antibody in 5-mC ELISA buffer was added into each well and incubated at 37°C for 1 h. Then, wells were washed with 5-mC ELISA buffer three times and an HRP developer was added to each well and incubated at room temperature for 30 min. The absorbance at 405 nm was measured using an ELISA plate reader. The 5-mC percentage for DNA samples was calculated using the logarithmic second-order regression equation at the standard curve that was constructed with negative and positive controls in the same experiment.

Gene Promoter CpG Island Methylation Quantification

ATG5 gene promoter methylation status was analyzed via Epitect Methyl-II PCR assay kit (Cat. No. 335002, QIAGEN, Hilden, Germany). Briefly, according to the manufacturer’s instructions and as described in previous studies (29, 30), genomic DNA was isolated from neonatal rat brain right hemisphere tissues and then DNA extracts were digested with EpiTect Methyl II DNA Restriction kit (Cat. No. 335452, QIAGEN, Hilden, Germany) in preparation for methylation analysis following the manufacturer’s protocol. The enzymes in the kit digest unmethylated and methylated DNA, respectively. After enzymatic cleavage, quantification of remaining input DNA in each enzyme reaction was conducted by real-time PCR with commercially available primer for the ATG5 gene provided by QIAGEN. The assay presents gene promoter methylation status as a percentage of methylated (M) and percentage of unmethylated (UM) fraction of input DNA.

Statistical Analysis

All statistical analysis was performed using GraphPad Prism 5 (GraphPad software). Data are expressed as means ± SE. Experimental number (n) represents neonates from different dams. Statistical significance (P < 0.05) was determined by analysis of variance followed by Newman–Keuls post hoc test or Student’s t test, where appropriate.

RESULTS

Maternal e-Cig Exposure Induces Growth Restriction and Programs a Brain Hypoxic-Ischemic Sensitive Phenotype in the Neonates

Rat pups were euthanized on P9. Body weight and brain weight were measured to determine the potential adverse effects of maternal e-cig exposure on neonatal development. As shown in Fig. 1, e-cig exposure during gestation significantly reduced body weight (Fig. 1A) and brain weight (Fig. 1B) in both male and female neonatal pups. However, e-cig exposure significantly increased the brain-to-body weight ratio (Fig. 1C).

Figure 1.

Maternal e-cigarette (e-cig) exposure induced an aberrant brain development in neonatal rat pups. Body weights (A) (n = 18–31/group), brain weights (B) (n = 5–10/group), and the brain-to-body weight ratio (C) (n = 5–10/group) of neonatal offspring were examined in both control and e-cig exposed groups at the age of postnatal day 9 (P9) of both male and female pups. Data were expressed as means ± SE of animal numbers (n) from each group. *P < 0.05, control vs. e-cigarette exposure.

Our previous studies have demonstrated that perinatal nicotine exposure programs a sensitive phenotype to neonatal HIE (13, 14). In the present study, we evaluated whether maternal e-cig exposure also induces similar effects in the neonatal brain. As shown in Fig. 2, there was a significant increase in hypoxic-ischemia (HI)-induced infarction size of the neonatal brain in males in the e-cig exposed group as compared with the control group. However, in female offspring, there was a tendency but not a statistically significant (P > 0.05) increase in HI-induced brain infarction size in e-cig exposed group as compared with the control group.

Figure 2.

Maternal e-cigarette (e-cig) exposure exaggerated hypoxia/ischemia (HI)-induced brain injury in neonatal offspring. The offspring pups on postnatal day 9 (P9) were conducted HI procedures as described in the materials and methods. Then, the HI-induced brain infarct size in both control and e-cig-exposed groups was measured by TTC staining 48 h after the HI procedures. Top: images represent the HI-induced brain infarct size determined by 2% TTC staining, in which the stained red colors in the brain represent viable areas and the white colors represent infarct areas. Bottom: summed infarct sizes, expressed as a percentage of the total brain weight. Data are represented as means ± SE of animal numbers (n = 8–11) from each group. *P < 0.05, control vs. e-cigarette exposure. TTC, 2,3,5-triphenyltetrazolium chloride monohydrate.

Maternal e-Cig Exposure Increases Reactive Oxygen Species and Enhances DNA Methylation in the Neonatal Brain

We evaluated the potential adverse effects of maternal e-cig exposure on ROS activity in the neonatal brain. As shown in Fig. 3A, e-cig exposure during gestation significantly increased ROS production in the neonatal brain of both male and female offspring as compared with the controls. In addition, we examined the key ROS-associated enzyme, NADPH oxidase 2 (NOX2) protein abundance in the neonatal brain. As shown in Fig. 3B, e-cig exposure significantly upregulated protein levels of NOX2 in neonatal brains isolated from both male and female offspring as compared with the controls.

Figure 3.

Maternal e-cigarette (e-cig) exposure increased reactive oxygen species in neonatal rat brains. The offspring neonatal pups were euthanized and brains were harvested on postnatal day 9 (P9). ROS level was measured in the right hemisphere brain tissue homogenates (A) and relative abundance of NOX2 protein was determined by Western blot (B). Data were expressed as means ± SE of animal numbers (n = 4 or 5) from each group. *P < 0.05, control vs. e-cigarette exposure. NOX2, NADPH oxidase 2; ROS, reactive oxygen species.

We further examined the effect of maternal e-cig exposure on DNA methylation profiles in the neonatal brain. As shown in Fig. 4A, e-cig exposure differentially upregulated DNMT1 protein abundances in male brains but downregulated its abundances in female brains as compared with the controls. As shown in Fig. 4B, protein abundances of DNMT3β in both male and female brains were significantly enhanced in e-cig exposed group as compared with the controls. As shown in Fig. 4C, maternal e-cig exposure significantly increased the level of global DNA methylation in male neonatal brains, however, there was a tendency but not statistical significance of increase in the global DNA methylation level in female brain as compared with the controls. Of interest, e-cig exposure specifically enhanced CpG methylation level in the ATG5 gene promoter region in male neonatal brains (Fig. 4D).

Figure 4.

Maternal e-cigarette (e-cig) exposure induced a heightened DNA methylation profile in neonatal rat brains. The offspring neonatal pups were euthanized and brains were harvested on postnatal day 9 (P9). Relative protein expression levels of DNMT1 and DNMT3b were determined by Western blotting analysis (A and B). Global DNA methylation levels (C) and CpG islands methylation levels at ATG5 promoter region (D) were also examined, respectively. Data were expressed as means ± SE of animal numbers (n = 4 or 5) from each group. *P < 0.05, control vs. e-cigarette exposure. ATG5, autophagy-related protein 5; DNMT1, DNA methyltransferase 1; DNMT3b, DNA methyltransferase 3 b.

Maternal e-Cig Exposure Suppresses Autophagy-Related Gene Expression and Sirtuin 1 Expression in the Neonatal Brain

Growing evidence has shown the crucial roles of autophagy in the pathology of diverse cardiovascular and neurological diseases (31–34). In the present study, we evaluated the impacts of maternal e-cigarette exposure on autophagic flux in neonatal brains. As shown in Fig. 5, the protein levels of ATG5 (Fig. 5A) and LC3β-1/LC3β-2 (Fig. 5B) in both male and female neonatal brains were significantly attenuated in e-cig-exposed group as compared with the controls. In addition, as shown in Fig. 6, the protein levels of sirtuin 1 (Sirt 1) in both male and female neonatal brains were also attenuated in e-cig exposed group as compared with the controls.

Figure 5.

Maternal e-cigarette (e-cig) exposure suppressed autophagic flux signaling in neonatal rat brain. The offspring neonatal pups were euthanized and brains were harvested on postnatal day 9 (P9). Relative protein expression levels of ATG5 (A) and LC3β-1/LC3β-2 (B) were determined via Western blotting analysis. Data were expressed as means ± SE of animal numbers (n = 4 or 5) from each group. *P < 0.05, control vs. e-cigarette exposure. ATG5, autophagy-related protein 5; LC3 β, microtubule-associated protein 1 light chain 3β.

Figure 6.

Maternal e-cigarette (e-cig) exposure suppressed sirtuin 1 protein expression in neonatal rat brain. The offspring neonatal pups were euthanized and brains were harvested on postnatal day 9 (P9). Relative protein expression levels of sirtuin 1 (SIRT1) were determined via Western blotting analysis. Data were expressed as means ± SE of animal numbers (n = 4 or 5) from each group. *P < 0.05, control vs. e-cigarette exposure.

Inhibition of ATG5 Enhances the HI-Induced Infarction in the Neonatal Brain

ATG5 is a key component in autophagic flux signaling, implicated in various physiological and pathological processes (35, 36). Given the findings that e-cig exposure attenuated ATG5 expression in the neonatal brain (Fig. 5A), we further tested whether the attenuated ATG5 plays a role in the setting of neonatal HIE insult. In this study, ATG5 siRNA was administered via intracerebral injection on P7, and HIE was conducted on P9 neonatal rat pups. As shown in Fig. 7A, siRNA administration significantly reduced ATG5 protein levels in the neonatal rat brain. Importantly, downregulation of ATG5 protein exaggerated HI-induced brain injury in the neonate (Fig. 7B).

Figure 7.

Knockdown of ATG5 enhanced hypoxia/ischemia (HI)-induced injury in neonatal rat brain. ATG5 siRNA and negative control were administered intracerebroventricularly on postnatal day 7. After 2 days of treatment, HI procedures were conducted in the neonates on postnatal day 9. After 2 days of intracerebroventricular treatment, the ATG5 protein levels in the neonatal brains were evaluated by Western blotting analysis (A). Data were expressed as means ± SE of animal numbers (n = 6) from each group. *P < 0.05, control vs. ATG5 siRNA. The brain infarct sizes were determined by TTC staining 48 h after HI procedure (B). Data were expressed as means ± SE of animal numbers (n = 8–11) from each group. *P < 0.05, control vs. ATG5 siRNA. ATG5, autophagy-related protein 5; TTC, 2,3,5-triphenyltetrazolium chloride monohydrate.

DISCUSSION

E-cigarette use has gained increasing popularity in the past 10 years, particularly among those in the reproductive age of young and pregnant women (37). Although it is well known that tobacco cigarette smoking during pregnancy has profound detrimental effects on both maternal and fetal health, there is limited data to demonstrate the detrimental effect of e-cig exposure during pregnancy on fetal and neonatal development. Furthermore, to our knowledge, up to now there is no evidence to demonstrate that e-cig exposure during pregnancy is safe for both maternal and fetal health. In the present study, we provide novel evidence that fetal e-cig exposure during pregnancy restricts offspring growth and brain development, leading to programming of brain hypoxic-ischemia sensitive phenotype in postnatal life. Specifically, our current data showed that 1) e-cig exposure decreased neonatal body weights and brain weights, but asymmetrically increased the brain-to-body weight ratio in the neonate; 2) e-cig exposure caused a sex-dependent increase in HI-induced brain injury in neonatal male offspring; 3) e-cig exposure altered DNA methylation patterns and specifically enhanced CpG methylation levels at ATG5 promoter in neonatal brains; and 4) e-cig exposure reprogrammed certain hypoxic-ischemia sensitive protein expression patterns, i.e., decreased autophagy-related gene ATG5/LC3β and Sirt 1 protein abundance but increased ROS-associated enzyme (NOX2) expression and ROS production in neonatal brains. These findings suggest that e-cig exposure during pregnancy is indeed an important risk factor and perinatal insult, which can cause offspring growth restriction and induce a fetal programming of neonatal brain hypoxic-ischemia sensitive phenotype via DNA methylation epigenetic regulation of autophagy signaling pathway.

The present finding that e-cig exposure during pregnancy decreased the body weight in both male and female neonates as compared with the controls, suggests that maternal e-cig exposure causes offspring growth restriction. Similarly, previous studies have shown that e-cig exposure during pregnancy decreases fetal and neonatal body weights (8, 38). Orzabal et al. (38) have further demonstrated that only e-cig with nicotine aerosols but not nicotine-free e-cig aerosols exposure during pregnancy significantly reduces offspring body weights, which suggests that e-cig exposure-induced offspring growth restriction may be regulated by nicotine. Indeed, our previous studies that maternal nicotine exposure decreased fetal and neonatal body weights in offspring (39), further suggest that nicotine is the key factor in e-cig aerosols contributing to the offspring growth restriction. However, although previous studies have demonstrated that nicotine alone can significantly affect fetal development, it remains largely unclear whether nicotine is the only component in e-cig vapor contributing to the detrimental effect on offspring. In fact, several previous studies have shown that chemicals other than nicotine in the e-cig liquids can alter offspring development and function (1, 40, 41). The major components of the BluPlus e-cig used in this study, in addition to nicotine, include vegetable glycerin, propylene glycol, and natural and artificial flavors. Future studies are needed to further investigate which component(s) of the e-cig in our model contribute to the e-cig-mediated detrimental effects on offspring.

It is well known that developmental growth restriction is associated with an increased risk for development of diseases later in life (42). Thus, growth restriction resulting from fetal e-cig exposure during pregnancy may increase risks for developing a dysfunctional phenotype in postnatal life. In the present study, we found that e-cig exposure decreased brain weights but increased the brain-to-body weight ratio in offspring neonates. This suggests that e-cig exposure induces asymmetric alteration of brain development. Of importance, the present findings that e-cig exposure during pregnancy gender-dependently enhanced HI-induced brain injury in neonatal male offspring, further suggest that maternal e-cig exposure can induce developmental neurocytotoxicity and fetal programming of a brain hypoxic-ischemia sensitive phenotype in postnatal life. Similarly, recent studies in different animal models have reported that maternal e-cigarette smoking exposure induces aberrant changes in the developing brain in offspring (1, 43), leading to development of offspring brain ischemic injury, cognitive dysfunction, and behavior disorder (44, 45). The brain ischemic sensitive phenotype seen here in the offspring has also been previously identified in offspring from cigarette smoke-exposed dams (46) and nicotine-exposed animals (13, 28, 35). Taken together, our present findings with previous studies (13, 44, 46) suggest that alteration of brain development and programming of brain ischemic sensitive phenotype in offspring are the most common detrimental outcomes from tobacco cigarette smoking, e-cig, and nicotine exposure during pregnancy.

There are at least three major epigenetic mechanisms including DNA methylation, histone modification, and noncoding RNA regulation, which significantly contribute to the developmental origins of disease later in life. Among these, DNA methylation may be a pivotal epigenetic mechanistic link between maternal e-cig exposure-mediated growth restriction and development of hypoxic-ischemic sensitive phenotype in offspring. Our previous studies have revealed that various maternal stresses including hypoxia, nicotine, cocaine, and diabetes induce aberrant DNA methylation patterns in offspring, which further affects specific genes expression patterns leading to the development of neuro- and cardiovascular dysfunctional phenotypes later in life (13, 25, 47, 48). Furthermore, it has been shown that both maternal cigarette smoking and e-cigarette exposure cause differential changes of DNA methylation patterns in the offspring in different tissues, including cord blood (49), placenta (50), brain (44), and lung (51). Consistent with these studies, in the present study, we also found that e-cig exposure increased global DNA methylation levels associated with enhanced Dnmt1 and Dnmt3β protein expression in male offspring brains. Interestingly, the present findings that e-cig exposure differentially decreased Dnmt1 but increased Dnmt3β protein abundance without significant change of DNA methylation in female offspring brains, suggest a gender dimorphism of DNA methylation patterns in response to fetal e-cig exposure. This gender difference of global DNA methylation patterns is well correlated to the gender difference of HI-induced brain injury in neonatal offspring. In addition to the global DNA hypermethylation, we also found that e-cig exposure significantly enhanced CpG methylation levels at the ATG5 gene promoter in neonatal male offspring brains. Collectively, our data suggest that fetal exposure to e-cig could result in gender-dependent DNA hypermethylation in the epigenome and specific gene promoter leading to the gender-dependent development of offspring brain ischemic sensitive phenotype.

Our present data show a CpG hypermethylation at the ATG5 gene promoter in the neonatal male offspring brain. This suggests that the changes of CpG methylation could lead to imprinting of ATG5 gene expression. Indeed, we found that e-cig exposure attenuated ATG5 protein abundance in the neonatal offspring brain as compared with the control. Furthermore, we also found that the protein levels of LC3β-1/2 were significantly attenuated by e-cig exposure as compared with the controls. These findings suggest that the autophagy signaling pathway is reprogrammed and downregulated in response to fetal e-cig exposure. Autophagy plays a key role in pathophysiological processes (31–34). The autophagy pathway is mediated by a group of autophagy-related genes (ATGs). ATG5 is a protein essential for the early stages of autophagosome formation. Recent studies suggest that ATG5 has a crucial role during early embryonic brain development. The deletion of ATG5 leads to abnormal brain development in the animal, resulting in behavioral defects caused by neurodegeneration (15). In addition, ATG5 deficiency leads to ROS accumulation, resulting in mitochondrial dysfunction and organ injury (16, 17). The present finding that e-cig exposure enhanced NOX2 expression and ROS production in offspring brains, suggests that the enhanced ROS may be mediated by ATG5 deficiency. Previous studies in our laboratory have demonstrated that maternal nicotine exposure-induced development of brain hypoxic-ischemic sensitive phenotype is associated with a downregulation of autophagy signaling (35). In the present study, we use a neonatal rat model to knockdown ATG5 via siRNA and our results clearly show that knockdown of the ATG5 gene significantly enhanced HI-induced brain injury, closely resembling the e-cigarette exposure-induced pathological outcomes in the neonatal HIE model. These observations suggest that e-cig exposure-mediated downregulation of ATG5 and autophagy signaling may play a key role in the development of brain ischemia-sensitive phenotype in postnatal life.

In the present study, a significant decrease in sirtuin (Sirt1) protein expression in the neonatal brains was observed in the e-cig-exposed group, which suggests that the Sirt1 repression may be a key molecular mechanism linking maternal e-cig exposure to its mediated detrimental outcomes in the offspring brain. It is well known that Sirt1 is an important regulator of epigenetics, which not only deacetylates histones to modulate chromatin function but also can deacetylate transcription factors to regulate the expression of target genes either positively or negatively. It has been suggested that Sirt1 is tightly linked to DNA methylation via deacetylation of Dnmt (52). Specifically, Sirt1 activation can deacetylate Dnmt, leading to increases in its methyltransferase activity and DNA methylation (52), which results in alteration of the target gene expression. Numerous studies suggest that autophagy and oxidative stress-associated genes are the major targets of Sirt 1 (53, 54). Lee et al. (53) have demonstrated that Sirt1 activity is necessary for the induction of starvation-induced autophagy and ATG5 is one of autophagy genes targeted by Sirt1. In addition, Sirt1 has been demonstrated to regulate oxidative stress via regulation of the NOX gene (54). Intriguingly, the inhibition of Sirt1 with pharmacological agents or siRNA leads to an elevation of ROS levels (55), suggesting a causal relationship between Sirt1 and ROS. Our recent studies in a gestational diabetes mellitus pregnant rat model have also demonstrated that Sirt1 can interact with DNA methylation, autophagy, and ROS signaling in which their interaction contributes to the aberrant development of heart ischemia-sensitive phenotype in offspring (25). Taken together, our current findings with previous studies suggest that the reduced Sirt1 expression may be a key epigenetic molecular regulator and network linker between fetal e-cig exposure and programming of brain hypoxic-ischemic sensitive phenotype in postnatal life.

Perspectives and Significance

In conclusion, our present study provides novel evidence that e-cigarette aerosol vaping during pregnancy is a prenatal insult that causes offspring growth restriction and reprogramming of autophagy-related genes via DNA methylation epigenetic mechanism, leading to the development of a brain hypoxic-ischemic sensitive phenotype in offspring. Specifically, based on our data and previous studies, we speculate that e-cig exposure causes Sirt1 gene repression in the developing brain, which differentially increases Dnmt expression and activity leading to gender-dependent increase in global DNA methylation and specific CpG methylation at the ATG5 gene promoter. The enhanced DNA methylation induces an epigenetic downregulation of ATG5 and autophagy-related gene expression. Finally, the e-cig exposure-mediated autophagy-related gene such as ATG5 deficiency induces oxidative stress. All of these aberrant fetal programming leads to subsequent development of a brain hypoxic-ischemic sensitive phenotype in postnatal life. These novel findings not only improve our understanding of the epigenetic molecular mechanisms underlying maternal e-cig exposure-induce fetal programming of brain developmental defects but also could provide novel therapeutic strategies to prevent or rescue the development of neonatal HIE and e-cig-related brain developmental disorder.

GRANTS

This work was supported by National Institutes of Health Grants HL135623, DA041492, and HD088039 (to D. Xiao). This project was partially supported by The Regents of the University of California, Research Grants Program Office, Tobacco-Related Disease Research Program (TRDRP) Grants T29IR0437 (to D. Xiao) and T30FT0936 (to B. Liu).

DISCLAIMERS

The funders had no role in experimental design, data collection and analysis, decision to publish, or preparation of the manuscript.

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

A.W., Y.L., and D.X. conceived and designed research; A.W., Y.L., Y.Z., Y.F., and B.L. performed experiments; A.W., Y.L., Y.Z., Y.F., and B.L. analyzed data; A.W., Y.L., Y.Z., B.L., and D.X. interpreted results of experiments; A.W., Y.L., Y.F., and D.X. prepared figures; A.W. and Y.L. drafted manuscript; A.W., Y.L., Y.Z., X.M.S., L.Z., and D.X. edited and revised manuscript; A.W., Y.L., Y.Z., B.L., X.M.S., L.Z., and D.X. approved final version of manuscript.