Enhanced antioxidant capacity prevents epitranscriptomic and cardiac alterations in adult offspring gestationally-exposed to ENM

Abstract

Maternal engineered nanomaterial (ENM) exposure during gestation has been associated with negative long-term effects on cardiovascular health in progeny. Here, we evaluate an epitranscriptomic mechanism that contributes to these chronic ramifications and whether overexpression of mitochondrial phospholipid hydroperoxide glutathione peroxidase (mPHGPx) can preserve cardiovascular function and bioenergetics in offspring following gestational nano-titanium dioxide (TiO₂) inhalation exposure. Wild-type (WT) and mPHGPx (Tg) dams were exposed to nano-TiO₂ aerosols with a mass concentration of 12.01 ± 0.50 mg/m³ starting from gestational day (GD) 5 for 360 mins/day for 6 nonconsecutive days over 8 days. Echocardiography was performed in pregnant dams, adult (11-week old) and fetal (GD 14) progeny. Mitochondrial function and global N⁶-methyladenosine (m⁶A) content were assessed in adult progeny. MPHGPx enzymatic function was further evaluated in adult progeny and m⁶A-RNA immunoprecipitation (RIP) was combined with RT-qPCR to evaluate m⁶A content in the 3′-UTR. Following gestational ENM exposure, global longitudinal strain (GLS) was 32% lower in WT adult offspring of WT dams, with preservation in WT offspring of Tg dams. MPHGPx activity was significantly reduced in WT offspring (29%) of WT ENM-exposed dams, but preserved in the progeny of Tg dams. M⁶A-RIP-qPCR for the SEC insertion sequence region of mPHGPx revealed hypermethylation in WT offspring from ENM-exposed WT dams, which was thwarted in the presence of the maternal transgene. Our findings implicate that m⁶A hypermethylation of mPHGPx may be culpable for diminished antioxidant capacity and resultant mitochondrial and cardiac deficits that persist into adulthood following gestational ENM inhalation exposure.

Introduction

While engineered nanomaterials (ENMs) and advanced materials, have enabled important innovations in biomedical engineering, the increased potential for their interactions with biological tissues warrants assessment of their ability to elicit undesirable toxicological effects (Bommarito Martin, and Fry 2017; Crispi, Miranda, and Gratacoset al. 2018; Pietroiusti et al. 2018). The application of nano-titanium dioxide (nano-TiO₂), a commonly utilized ENM, escalated by 300% from 2011–2013 in consumer products (Vance et al. 2015). The primary route of exposure is inhalation, which allows particles to enter the bloodstream and systemically impact the body (Stebounova et al. 2012). The size of these particles (≥1 dimension that is ≤100 nm in diameter), plays a critical role in their ability to access systemic circulation and promote cardiovascular damage (Robichaud et al. 2009; Kessler 2011; Stebounova et al. 2012, Pietroiusti et al. 2018; Kunovac et al. 2020). Nano-TiO₂ has been implicated in alterations to cardiovascular hemodynamics and mitochondrial dysregulation (Hathaway et al. 2017; Nichols et al. 2018; Hathaway et al. 2019a). Exposure to ENMs are of particular concern for vulnerable populations such as the developing fetus (Bommarito et al. 2017; Ferrari, Carugno, and Bollati 2019). Although research relating to particulate matter exposure during gestational development and the associated detriments are receiving more attention, therapeutic interventions are limited and the identification of mechanisms responsible for long term effects remain undefined.

Initial studies from our laboratory have determined that maternal ENM inhalation exposure during gestation elicits cardiac contractile dysfunction and bioenergetic disruption in the developing fetus (Kunovac et al. 2019), which are sustained into adulthood (Hathaway et al. 2017; Kunovac et al. 2019). At the fetal stage, the functional deficits were attributed to an increased production of reactive oxygen species (ROS) (marked by hydrogen peroxide (H₂O₂) levels) and downregulation of mitochondria phospholipid hydroperoxide glutathione peroxidase (mPHGPx) (Kunovac et al. 2019), an antioxidant enzyme that protects cells from oxidative stress. Because mPHGPx scavenges hydroperoxides, limiting diffusion of H₂O₂ from the mitochondrion into other cellular regions, its expression is critical for decreasing ROS influence on nuclear genome regulation. MPHGPx has the ability to mitigate cardiac dysfunction and bioenergetic dysregulation in several contexts (Hollander et al. 2003; Dabkowski et al. 2008; Baseler et al. 2013), including acute nano-TiO₂ exposure (Nichols et al. 2018). However, whether enhanced antioxidant capacity can provide a protective role to fetal and adult offspring following maternal ENM inhalation exposure during gestation remains unexplored.

In a previous study from our laboratory, we reported mitochondrial and cardiac contractile decrements in the fetal stage that were sustained into adulthood following maternal ENM inhalation exposure (Kunovac et al. 2019). Though the mechanism by which these detriments are sustained is unclear, epitranscriptomic modifications resulting from maternal ENM exposure may be contributing to the persistent cardiac mitochondrial dysfunction observed. Of interest was our observation of increased ROS and concomitant diminution of mPHGPx protein expression at the fetal stage which may have been associated with epigenetic reprogramming that occurs following particulate exposure (Bommarito et al. 2017; Janssen et al. 2017; Kietzmann et al. 2017; Tanwar et al. 2017; Stapleton et al. 2018; Ferrari et al. 2019; Kunovac et al. 2019). Specifically, epitranscriptomic mechanisms may be responsible for degradation of mRNA and/or functional deficits in proteins (Engel et al. 2018; Min et al. 2018; Leonardi et al. 2019). N⁶-methyladenosine (m⁶A) is the most prevalent post-transcriptional mRNA modification in eukaryotes and has been associated with various pathologies (Ross-Innes et al. 2012; Engel et al. 2018; Zhong et al. 2018; Castellanos-Rubio et al. 2019; Cayir et al. 2019; Linder and Jaffrey 2019). Elevated stress levels have the ability to alter m⁶A machinery and therefore, modulate transcript stability by adding and removing methylation marks (Shi et al. 2017; Engel et al. 2018; Zhong et al. 2018; Leonardi et al. 2019; Berulava et al. 2020). M⁶A sites are known to be enriched near stop codons and in 3′-untranslated regions (UTR) of mRNA (Wan et al. 2015), thereby influencing translation (Meyer 2019). Investigation of m⁶A sites in crucial coding regions of nuclear genome-encoded mitochondrial mRNAs that are pivotal for antioxidant defense maintenance is of particular interest, as it may provide insight into mechanisms that result in the longitudinal outcomes of maternal ENM exposure during gestation.

We hypothesize that altered m⁶A status of the 3′-UTR of mPHGPx in the hearts of adult offspring contributes to mitochondrial and cardiac functional deficits and that overexpression of mPHGPx (in the mother or pup) can preserves these functional alterations following maternal ENM inhalation exposure. Through the use of differential breeding strategies, we were able to determine that the maternal transgene is protective for the fetus and the benefit is sustained in adult offspring, regardless of whether the progeny also possessed the transgene. Following gestational ENM inhalation exposure, mPHGPx enzymatic activity was entirely preserved only in offspring whose dams were mPHGPx transgenic. M⁶A methylation was highly enriched in the 3′-UTR of mPHGPx in wild-type (WT) offspring exposed in utero to nano-TiO₂, but maternal mPHGPx overexpression prevented this modification. Our findings suggest that enhancing antioxidant defense in the pregnant dam may provide offspring with the most protection from epitranscriptomic remodeling and subsequent long-term cardiovascular repercussions that arise following gestational nano-TiO₂ inhalation exposure.

Materials and methods

Animal model

The West Virginia University Animal Care and Use Committee approved all animal studies which conform to the most current National Institutes of Health (NIH) Guidelines for the Care and Use of Laboratory Animals (8^th edition) manual. Male and female transgenic (mPHGPx (+/+)) and wild-type (mPHGPx (−/−)) mice with a FVB/NJ genetic background were housed in the West Virginia University Health Sciences Center Animal Facility. All animals were given access to a rodent diet and water ad libitum. The mPHGPx transgenic mouse model has been previously described by our laboratory (Dabkowski et al. 2008; Baseler et al. 2013). Briefly, to achieve overexpression of mPHGPx, the gene was placed under the control of the cytomegalovirus promoter, which provides robust transgene expression. Transgene screening was performed on tail clip DNA of 3-week old mice using a qPCR approach in which we probed for mPHGPx using a fluorometric probe (Product no. Mm00515041_m1, Applied Biosystems, Foster City, CA). Increased mPHGPx protein content and improved scavenging ability for phosphatidylcholine hydroperoxide (PCOOH) and hydrogen peroxide (H₂O₂) in the mPHGPx transgenic mice indicated a functionally active enzyme (Dabkowski et al. 2008). MPHGPx overexpression as an antioxidant protective strategy was tested in multiple contexts utilizing complimentary breeding strategies (Figure 1(A,B)). Because mPHGPx mice are heterozygous transgenic, the transgene status of the donor parent (paternal vs. maternal) impacts the nature of the protective strategy provided to the offspring. The presence of the paternal mPHGPx transgene (Paternal mPHGPx Maternal Exposure (PPME)) enabled a portion of the litter (~50%) to be mPHGPx transgenic positive (Figure 1(A), right panel) and thus presented the opportunity to determine whether fetal mPHGPx transgene presence is protective in utero and adult offspring. The presence of the maternal mPHGPx transgene (Maternal mPHGPx maternal exposure (MPME)) enabled a portion of the litter (~50%) to be mPHGPx transgenic (Tg) as well, but with the added benefit of the maternal mPHGPx transgene (Figure 1(A), left panel). This breeding strategy allowed us to delineate whether mPHGPx transgene presence in the environment provided by the dam is beneficial to the non-transgenic progeny or of additional benefit to transgenic progeny. Pregnancy was verified by identifying the vaginal plug. Approximately 5 days after identification of the vaginal plug (gestational day 5 (GD 5)), the pregnant dams were designated into either sham or nano-TiO₂ exposure groups. Echocardiographic assessments were performed on adult offspring (11 weeks old), as well as pregnant dams (12 weeks old) and fetal-stage progeny (GD 14). For the adult offspring study, pups remained with the singly-housed dams until weaning age (19 days) at which point they were housed with littermates and separated by sex, with no more than 5 animals per cage. Adult progeny was euthanized at ~11 weeks of age and hearts were excised for further analyses. From each adult litter, one animal of each genotype (WT/Tg) was randomly selected and designated N = 1, alternating selection of male and female offspring to control for sex as a biological variable. All offspring were genotyped using the qPCR method described above. Pregnant dams were euthanized one day following echocardiographic imaging (GD 15), the pups were removed and tail clips were collected for genotyping. Adult offspring and pregnant dams were euthanized through sedation with 5% isoflurane and subsequent cervical dislocation. A timeline of the study is provided in Figure 1(C).

Figure 1.

MPHGPx mouse breeding strategy. (A) Schematic of mPHGPx breeding strategies with transgene effect on offspring from maternal (left) or paternal (right) transgene; (B) Schematic of exposure paradigm that was implemented for each group, representing all groups that were utilized in the study; (C) A timeline of the study. MPHGPx: mitochondrial phospholipid hydroperoxide glutathione peroxidase.

Engineered nanomaterial inhalation exposure

The ENM inhalation exposure paradigm utilized in the current study has been previously described (Hathaway et al. 2019a). Nano-TiO₂ P25 powder containing anatase (80%) and rutile (20%) TiO₂ was purchased from Evonik (Aeroxide TiO₂, Parsipanny, NJ) and prepared by drying, sieving, and storing (Nurkiewicz et al. 2008; Knuckles et al. 2012; Hathaway et al. 2019a). The primary particle characteristics have been previously reported including the size (21 nm), the specific surface area (48.08 m²/g), and the Zeta potential (−56.6 mV) (Nurkiewicz et al. 2008; Sager, Kommineni, and Castranova 2008; Nichols et al. 2018). Nano-TiO₂ was aerosolized with a high-pressure acoustical generator (HPAG; IEStechno, Morgantown, WV), and has been detailed in previous studies (Hathaway et al. 2019a) involving rodent inhalation exposure. Figure 2 provides data outlining the aerosol characterization of nano-TiO₂. In order to recapitulate a lung burden typically seen in a manufacturing setting exposure (Kunovac et al. 2019), a target aerosol mass concentration of 12 mg/m³ of engineered nano-TiO₂ was chosen, for 360 min per day for 6 nonconsecutive days over an 8-day period, with a whole-body exposure chamber. This level was based on human equivalent alveolar doses during pregnancy and is explained in more detail later in the manuscript. The real-time TiO₂ aerosol mass concentration readings (pDR-1500; Thermo Environmental Instruments Inc., Franklin, MA) over a typical exposure day were sampled from an exposure chamber (Figure 2(A)) and verified by gravimetric measurements during each exposure. Final gravimetric measurements indicated a daily 360-min equivalent average mass concentration of 12.01 ± 0.50 mg/m³. A high-resolution electrical low-pressure impactor (ELPI+; Dekati, Tampere, Finland), a scanning particle mobility sizer (SMPS 3938; TSI Inc., St. Paul, MN), and a Nano Micro-Orifice Uniform Deposit Impactor (MOUDI 115 R; MSP Corp, Shoreview, MN) were used to measure the size of the nano-TiO₂ aerosols. A log-normal fit of the data from the ELPI + indicated a geometric count median diameter (CMD) of 0.163 μm with a geometric standard deviation (GSD) of 1.77 (Figure 2(B)). A log-normal fit of the data from the SMPS indicated a CMD of 0.190 μm with a GSD of 1.97 (Figure 2(C)). A log-normal fit of the data from the MOUDI indicated a mass median aerodynamic diameter (MMAD) of 0.968 μm with a GSD of 2.56 (Figure 2(D)). Transmission and scanning electron micrographs (TEM and SEM) of nano-TiO₂ aerosolized particles, sampled from the exposure chamber have been previously published (Kunovac et al. 2019). The dose required to match the appropriate lung deposition was calculated based on the previously described mouse methodology modified to reflect alveolar deposition only (Nichols et al. 2018). The formula D = F × V × C × T, where F is the alveolar deposition fraction (4.40%) ((Multiple Path Particle Dosimetry Model (MPPD v3.04)) (Anjilvel and Asgharian 1995), V is the minute ventilation based on body weight (36.4 ml) (Bide, Armour, and Yee 1997), C is the mass concentration (12.01 mg/m³) and T is the exposure duration (6 h), was employed (Nurkiewicz et al. 2008; Stapleton et al. 2012). This resulted in a daily deposited nano-TiO₂ alveolar dose of 6.92 μg (total six exposure dose = 41.55 μg). When this alveolar burden is normalized with respect to alveolar surface area (mouse_sa = 0.05 m³, human_sa = 102 m³) and matched to human breathing the same aerosol, it corresponds to a human alveolar deposition of 84.76 mg. This is equivalent to an approximately 79-day exposure at the NIOSH recommended exposure limit (REL) of 2.4 mg/m³ for fine TiO₂ established in 2011 (Department of Health and Human Services, 2011) (F = 12.24% (MMPD), V = 7.6 L). Since factory workers are exposed to airborneTiO₂ levels of ~0.65 mg/m³ (Pelclova et al. 2015) and the human gestational period is 9 months, we feel that our exposure concentration levels are highly relevant to the female worker population. To maintain a comfortable humidity during the exposure, bedding material was soaked in water and placed in the exposure chamber. Control animals (sham) were exposed to HEPA filtered air, rather than nano-TiO₂, in a designated chamber with similar chamber conditions. The final exposure was administered 48 h prior to sacrifice and tissue harvesting.

Figure 2.

Maternal nano-TiO₂ inhalation exposure characteristics. (A) Real-time (black line) aerosol mass concentration measurements with the red line indicating the target concentration (12 mg/m³) of a typical 360 min maternal nano-TiO₂ inhalation exposure; (B) Count size distribution of the nano-TiO₂ aerosols measured with a high resolution electric low-pressure impactor (ELPI+). The red line designates the log normal distribution obtained with the log probability plot method (CMD = 0.163 μm, with a GSD = 1.77); (C) Count size distribution of nano-TiO₂ aerosols measured with a scanning mobility particle sizer (SMPS) with the red line representing a log normal fit of the data (CMD = 0.190 μm, GSD = 1.97); (D) Mass size distribution of nano-TiO₂ aerosols measured with a nano micro-orifice uniform deposit impactor (MOUDI). The red line represents a log normal fit of the data indicating a mass median aerodynamic diameter (MMAD) of 0.968 μm and a GSD of 2.56. CMD: count median diameter; GSD: geometric standard deviation.

Echocardiography

Cardiac contractile function was assessed in adult offspring (11 weeks old) using the Vevo2100 High-frequency Ultrasound system (Visual Sonics, Toronto, Canada) to obtain Motion mode (M-mode) images as previously described (Kunovac et al. 2019). Anesthesia was induced using inhalant isoflurane, which was then maintained at ~1.5% in order to achieve a physiologically relevant heart rate. Images were acquired with a linear array transducer at 32–40 MHz, with a frame rate of 233–401 frames/s. M-mode images were also obtained for pregnant dams (12 weeks old) and their fetal progeny (GD 14). For fetal echocardiographic imaging, pups in the right and left uterine horns were imaged sequentially, which allowed us to correlate the assessed cardiac function with the genotype of that specific pup. For adult offspring analyses, one animal of each genotype (WT/Tg) from each exposed (sham/nano-TiO₂) dam (PPME (WT mom) or MPME (Tg mom) was selected at random. Measurements were calculated over three cardiac cycles and averaged.

Speckle-tracking-based strain evaluations were performed using parasternal long and short axis B-mode images as previously described by our laboratory (Nichols et al. 2015; Shepherd et al. 2016; Hathaway et al. 2017; Kunovac et al. 2019). A speckle-tracking algorithm in Visual Sonics VevoStrain software (Toronto, Canada) was employed to trace the endocardium and epicardium walls. Data were analyzed for three cardiac cycles using time-to-peak analysis for curvilinear data. Measures of systolic strain including deformation, strain rate, displacement, and velocity were obtained. For adult progeny, speckle-tracking was performed on both long- and short-axis images. For fetal progeny, speckle-tracking was performed on short-axis images. All echocardiographic measurements were acquired by one analyst blinded to the animal exposure group and genotype.

Electron transport chain (ETC) complex activities

ETC Complex activities (I, III, IV, and V) were measured in hearts of adult offspring as previously described (Barrientos, Fontanesi, and Diaz 2009; Baseler et al. 2013; Hathaway et al. 2017). Whole tissue was homogenized using the Polytron PowerGen 500 S1 tissue homogenizer (Fisher Scientific, Hampton, NH) in NP-40 buffer (20 mM Tris, 137 mM NaCl, 10% Glycerol, 1% Triton x100, 2 mM EDTA) (Hathaway et al. 2019b). Samples were centrifuged for 10 min at 10 000 xg (4 °C) and the supernatant was used for the activity assays. The protein homogenates were used to measure activities of ETC complexes I, III, IV, and V (ATP synthase). ETC complex I and III activities were determined by measuring the reduction of decylubiquinone (I) and cytochrome c (III) in cardiac protein lysate of adult offspring. ETC complex IV activity was determined by measuring the oxidation of reduced cytochrome c, while complex V activity was determined by measuring oligomycin-sensitive ATPase activity through pyruvate kinase and phosphoenolpyruvate in cardiac protein lysate of adult offspring. The Molecular Devices Flex Station 3 Multi-Mode microplate reader (Sunnyvale, CA) was used to measure all assays spectrophotometrically. Protein content was normalized using the Bradford method, with bovine serum albumin protein assay standards. Final values were expressed as Unit/nanogram (I-IV) or milligram (V) of protein, where Unit = nanomoles of substrate oxidized (minute⁻¹).

Total glutathione peroxidase (GPx) activity

Total GPx activity was measured using a GPx Activity Kit from Cayman Chemical (item no. 703102, Ann Arbor, MI). The Cayman Chemical GPx Activity Kit determines all glutathione-dependent peroxidase activity indirectly through a coupled reaction with glutathione reductase (GR). The cumene hydroperoxide in this assay is reduced by GPx producing oxidized glutathione (GSSG), which can then be recycled to its reduced state by GR and NADPH. The oxidation of NADPH results in a decrease in absorbance at 340 nm that is proportional to the GPx activity when it is the rate-limiting factor. The experiment was carried out using the manufacturer’s protocol, with minor modifications: Heart lysate (in NP-40) samples were diluted with sample buffer and 20 μl of the diluted heart lysate was used for the assay. Absorbance was read once every minute for 6 min at 340 nm using the Molecular Devices Flex Station 3 Multi-Mode microplate reader. Absorbance values were retrospectively normalized to protein content with the Bradford assay (Bradford 1976).

MPHGPx activity

To determine mitochondrial phospholipid hydroperoxide glutathione peroxidase (GPx4) activity, phosphatidylcholine hydroperoxide (PCOOH), a specific substrate of mPHGPx, was synthesized as previously described (Stolwijk et al. 2020). l-α-Phosphatidylcholine Type III/S (PC) (item no. P3782; Sigma-Aldrich, St. Louis, MO) was slowly added to a Tris/Base buffer (0.2 M, pH 8.8, 3 mM sodium deoxycholate (item no. D6750, Sigma-Aldrich)) while stirring at medium-high speed. Once a cloudy emulsion of small droplets was visible, additional Tris/Base buffer was introduced with continuous stirring until the solution was clear. A volume of soybean lipoxidase Type V (item no. L6632, Sigma-Aldrich) was added that equaled 250 000 U to initiate the synthesis reaction followed by continuous stirring for 1 h. A Sep-Pak C-18 cartridge (item no. 020515, Waters, Milford, MA) was activated with methanol and equilibrated with ddH₂O, then used to purify the PCOOH with a 30 ml glass syringe (item no. Z314374, Sigma-Aldrich). The cartridge was washed with water to remove water-soluble substances and the PCOOH was eluted from the C18 resin with methanol. The concentration of PCOOH was determined at A₂₃₄ using the NanoDrop ND-100 (Thermo Fisher Scientific, Waltham, MA) and diluted to the desired concentration with methanol. The PCOOH was immediately used as the substrate (in place of cumene hydroperoxide) with the Glutathione Peroxidase Assay Kit (Cayman) as described above, with absorbance readings every 30 s for 6 min at 340 nm with the Molecular Devices Flex Station 3 Multi-Mode microplate reader. All samples were blanked with background wells (no sample) and normalized to protein content.

Mitochondrial isolation

Adult progeny were sacrificed at 11 weeks of age, and hearts were excised through a midsagittal cut in the thoracic cavity. Isolation of mitochondrial subpopulations was achieved through differential centrifugation as previously described (Palmer, Tandler, and Hoppel 1977; Baseler et al. 2013). The two subpopulations were combined and utilized as a total mitochondrial fraction for subsequent analyses. KME buffer (100 mM KCl, 50 mM MOPS and 0.5 mM EGTA pH 7.4) was used to resuspend isolated mitochondria. Protein concentrations were determined using the Bradford method with bovine serum albumin as a standard (Bradford 1976).

Hydrogen peroxide (H₂O₂) production

Hydrogen peroxide (H₂O₂) production was evaluated in total tissue lysate and isolated mitochondria from adult offspring in order to determine mitochondrial ROS production (Starkov 2010; Nichols et al. 2018; Kunovac et al. 2019). The Invitrogen™ Amplex™ Red Hydrogen Peroxide Assay Kit (product no. A2218; Thermo Fisher) was utilized per the manufacturer’s instructions. In the presence of horseradish peroxidase (HRP), the Amplex™ Red reagent produces a red fluorescent oxidation product when the reagent reacts with H₂O₂. The oxidation product, resorufin, was detected spectrophotometrically at 560 nm and data were normalized to protein content.

Global M⁶A methylation

Total RNA was purified from heart samples that were digested with QIAzol lysis reagent (item no. 79306, Qiagen, Hilden, Germany) from one adult offspring of each genotype (WT/Tg) from each dam (PPME/MPME) from each exposure group (sham/nano-TiO₂) using the miRNeasy Mini Kit (item no. 217004, Qiagen). Global m⁶A methylation was quantified in total RNA using the EpiQuik™ m⁶A RNA Methylation Quantification Kit (item no. P-9005; EpiGentek, Farmingdale, NY) per manufacturer’s instructions. Briefly, total RNA samples were diluted to a standard concentration of 200 ng with IDTE buffer. Dilution of the provided positive control allowed for the generation of a standard curve. RNA was bound to the assay wells through incubation with the binding solution. An optimized m⁶A capture antibody was added to the wells, incubated, and washed, followed by the addition of the detection antibody and enhancer solution. The color developing solution was then introduced and incubated for 2 min, until the color changed to blue, at which point the stop solution was added. Absorbance was read at 450 nm using the Molecular Devices Flex Station 3 Multi-Mode microplate reader. A standard curve was used to determine global m⁶A levels in each sample.

Immunoprecipitation (m⁶A-RIP)

Total RNA was purified from whole heart samples from PPME and MPME WT adult offspring from each exposure group (sham/nano-TiO₂) using the miRNeasy Kit (Qiagen) described above.

RNA fragmentation

Before the immunoprecipitation was carried out, the RNA samples were fragmented using the NEBNext Magnesium RNA Fragmentation kit (item no. E6150s; New England Biolabs) (Zeng et al. 2018; Barros-Silva et al. 2020). To determine the necessary length of incubation to achieve desired fragmentation (~130 nt), samples were incubated for 1–10 min at 94 °C. The 10 samples were analyzed with the Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA) to determine the size distribution of the fragments.

M⁶A-RIP

The EpiMark N6-Methyladenosine Enrichment Kit (item no. E1610S; New England Biolabs, Ipswich, MA) was used to enrich m⁶A modified RNA per manufacturer’s instructions, with minor modifications. RNA samples were fragmented at 94 °C for 3 min (Barros-Silva et al. 2020). Dynabeads™ Protein G magnetic beads from Thermo Fisher Scientific (product no. 10003 D) were washed and resuspended in Reaction Buffer. An anti-N6-Methyladenosine (m⁶A) rabbit monoclonal antibody (product no. 56593, Cell Signaling Technology, Danvers, MA) was bound to the beads, followed by the addition of the fragmented total RNA (3.7 μg) and RNasin ® Plus RNase Inhibitor (product no. N2611; Promega, Madison, WI). The normal Rabbit IgG control (product no. 2729, Cell Signaling Technology), an unconjugated rabbit polyclonal antibody, was used as a nonspecific IgG control for the immunoprecipitation. After the immunoprecipitation, RNA was eluted with Buffer RLT (product no. 79216; Qiagen) then cleaned and concentrated using Dynabeads™ MyOne™ Silane (product no. 37002 D; Life Technologies, Carlsbad, CA). The RNA samples from both input and immunoprecipitated samples were examined by RT-qPCR to quantify m⁶A enrichment.

RT-qPCR

WT adult offspring heart samples (from WT/Tg and sham/nano-TiO₂ exposed dams) were prepared for downstream analysis by reverse-transcribing the input and immunoprecipitated RNA using the Applied Biosystems High-Capacity cDNA Reverse Transcription Kit (product no. 4368814; Thermo Fisher Scientific) per manufacturer’s protocol. RT-qPCR was utilized to evaluate m⁶A enrichment of the mPHGPx 3′-UTR, specifically the region containing the crucial selenocysteine (Sec) insertion sequence (SECIS) (Ingold et al. 2018; Castellanos-Rubio et al. 2019). Primers for mPHGPx and a control, GAPDH, were designed using Primer3Plus (Baseler et al. 2011) and the sequences are listed in Table S1 (Additional file 1). Quantification was achieved with SYBR^® Green master mix (product no. A25742, Thermo Fisher Scientific) using the Applied Biosystems 7500 Fast Real-Time PCR system. Samples were run in triplicate. The enrichment fold was calculated using the ΔCt value of the IP normalized to the ΔCt of the input, relative to the dilution factor (20% input; dilution factor = 5), and normalized to GAPDH, as illustrated below (Marmisolle, Garcia, and Reyes 2018).

$$\begin{gathered} \mathrm{\Delta }Ct\left[NormalizedIP\right]=Ct\left[IP\right] \\ -\left(Ct\left[Input\right]-Log2\left(DilutionFactor\right)\right) \\ \mathrm{\Delta }\mathrm{\Delta }Ct=\mathrm{\Delta }Ct\left[NormalizedI{P}_{(mPHGPx)}\right] \\ -\mathrm{\Delta }Ct\left[NormalizedI{P}_{(GAPDH)}\right] \\ Foldenrichment=2^{(-\mathrm{\Delta }\mathrm{\Delta }Ct)} \end{gathered}$$

Statistics

All statistical analyses were performed using GraphPad Prism Software Version 8 for Windows (GraphPad Software, La Jolla CA). For adult offspring analyses, from each MPME or PPME cohort, one animal of each genotype (WT/Tg) from a given exposure (sham and nano-TiO₂) was randomly selected and designated N = 1. For fetal analyses, each pregnant dam that was exposed (sham/nano-TiO₂) was considered one observation, with each genetic group (WT/Tg) considered as N = 1. Data were analyzed using a two-way analysis of variance (ANOVA) method to evaluate mPHGPx presence and nano-TiO₂ maternal exposure. Normality was determined using the D’Agostino-Pearson test. Tukey’s multiple comparison method was performed following the two-way ANOVA to evaluate effects within the PPME and MPME groups. Statistical difference was defined by p ≤ 0.05. The presence of a letter above a specific group denotes statistical significance between those two groups as determined by Tukey’s multiple comparisons test. For functional assays performed using adult hearts, a two-tailed Student’s t-test was also employed to evaluate whether maternal nano-TiO₂ inhalation exposure during gestation elicited change between offspring of the same genotype. The presence of a dagger (†) above a given group indicates statistical significance between WT Sham and WT TiO₂ or Tg Sham and Tg TiO₂ of the same maternal genetic group (MPME or PPME). For qPCR data, a one-way ANOVA was employed followed by Tukey’s multiple comparisons test to determine significant differences in m⁶A fold enrichment across WT offspring groups. All data are presented as the mean ± the standard error of the mean (SEM).

Results

Cardiac function

Deleterious changes in cardiac structure and function have previously been identified in offspring following maternal nano-TiO₂ inhalation exposure during gestation (Hathaway et al. 2017; Kunovac et al. 2019). Conventional M-mode evaluations determined that adult offspring of WT dams (PPME) who were gestationally exposed to nano-TiO₂ presented with higher systolic diameter and volume, as well as lower ejection fraction and fractional shortening than those whose dams were sham-exposed (Table 1). Additionally, adult offspring whose dams were mPHGPx Tg (MPME) overall had significantly lower diastolic diameter and volume, stroke volume, and cardiac outputs than PPME offspring regardless of exposure (Table 1). Adult offspring heart rates were unchanged regardless of exposure or genotype (Additional file 1: Table S2).

Table 1.

M-mode echocardiography for adult progeny.

Parameter	PPME WT Shama	PPME WT TiO2b	PPME Tg Shamc	PPME Tg TiO2d	MPME WT Shame	MPME WT TiO2f	MPME Tg Shamg	MPME Tg TiO2h
Diameter; s (mm)	0.61 ± 0.05	0.85 ± 0.08 a,d	0.66 ± 0.07	0.51 ± 0.06	0.63 ± 0.05	0.60 ± 0.02	0.57 ± 0.02	0.72 ± 0.06
Diameter; d (mm)	2.12 ± 0.04	2.27 ± 0.08	2.15 ± 0.09	2.07 ± 0.07	1.98 ± 0.08	1.98 ± 0.07	1.88 ± 0.06	2.07 ± 0.07
Volume; s (uL)	0.58 ±0.13	1.47 ± 0.35 a,c,d	0.73 ± 0.19	0.38 ± 0.11	0.60 ± 0.11	0.56 ± 0.02	0.46 ± 0.05	0.94 ± 0.24
Volume;d (uL)	14. 78 ± 0.61	17.85 ± 1.66	15.63 ± 1.56	14.09 ± 1.25	12.61 ± 1.31	12.65 ± 1.11	11.00 ± 0.92	14.21 ± 1.31
Stroke volume (uL)	14. 19 ± 0.63	16.38 ± 1.34	14.90 ± 1.43	13.71 ±1.16	12.01 ± 1.21	12.08 ± 1.12	10.54 ± 0.87	13.27 ± 1.50
Ejection fraction (%)	96.02 ± 0.94	92.22 ± 1.29 a,d	95.55 ± 0.91	97.48 ± 0.59	95.34 ± 0.41	95.43 ± 0.53	95.82 ± 0.13	93.35 ± 2.30
Fractional shortening (%)	71.25 ±2.37	62.99 ± 2.24 a,d	69.84 ± 2.37	75.72 ± 2.37	68.64 ± 1.24	69.64 ± 1.92	69.52 ± 0.35	65.40 ± 4.27
Cardiac output (mL/min)	10.08 ± 0.56	10.32 ± 0.77	9.67 ± 1.07	9.06 ±0.81	7.60 ± 1.44	8.00 ± 0.71	6.80 ± 0.78	9.19 ± 1.20
Relative wall thickness	1.23 ± 0.06	1.20 ± 0.06	1.17 ± 0.12	1.36 ± 0.06	1.18 ± 0.06	1.27 ± 0.06	1.35 ± 0.05	1.15 ± 0.10

M-mode echocardiography was performed in adult offspring (11 weeks of age) following maternal inhalation exposure. Bold text indicates significant data defined by p ≤ 0.05. A two-way ANOVA determined statistical difference between PPME and MPME denoted with underlined values for that parameter. A letter next to a given group denotes statistical significance between the marked group and the group represented by each letter based on a Tukey’s multiple-comparisons test. All data are presented as the mean ± the standard error of the mean (SEM). PPME WT Sham, n = 7; PPME WT TiO₂, n = 6; PPME Tg Sham, n = 6; PPME Tg TiO₂, n = 5; MPME WT Sham, n = 4; MPME WT TiO₂, n = 5; MPME Tg Sham, n = 5; MPME Tg TiO₂, n = 4.

PPME: paternal mPHGPx maternal exposure; MPME: maternal mPHGPx maternal exposure; WT Sham: wild-type offspring whose dam was exposed to control air; Tg Sham: mPHGPx transgenic offspring whose dam was exposed to control air; WT TiO₂: wild-type offspring whose dam was exposed to nano-TiO₂; Tg TiO_2: mPHGPx transgenic offspring whose dam was exposed to nano-TiO₂; s: systolic; d: diastolic.

When examining the dams, WT nano-TiO₂-exposed mice had significantly decreased diastolic diameter and volume when compared to WT sham-exposed dams, as well as reduced pump function indicated by stroke volume and cardiac output reductions (Additional file 1: Table S3). Furthermore, relative wall thickness was significantly increased following ENM exposure in WT dams, but not in the mPHGPx Tg dams (Additional file 1: Table S3). Maternal heart rate remained unchanged between all exposure groups (Additional file 1: Table S4). At the fetal stage, in order to correlate the assessed cardiac function with the genotype of that specific pup, pups in the right and left uterine horns were sequentially imaged (Figure 3(A)). A representative trace of the fetal pup’s left ventricle (LV) in M-mode is shown in Figure 3(B). A group-dependent effect was found, with the MPME fetal offspring demonstrating higher diastolic diameter and volume, as well as higher stroke volume and cardiac output than the PPME offspring (Additional file 1: Table S5). MPME fetal offspring also had significantly higher heart rates (Additional file 1: Table S2). Decreased function in the PPME WT offspring following gestational ENM exposure confirms previous findings (Kunovac et al. 2019) and substantiates the susceptibility of WT offspring to cardiovascular changes due to maternal nano-TiO₂ exposure, which may be a result of limited antioxidant capacity.

Figure 3.

Assessment of fetal pup cardiac function in utero. (A) Graphical depiction of the uterine horn and the pup identification system; (B) Representative M-mode echocardiographic scan of a fetal pup with the short-axis left ventricular trace (blue); (C–D) Representative B-mode scan of a fetal pup with the left and right ventricles identified by blue arrows at end diastole (C) and end systole (D); (E–F) Representative B-mode scan in the short-axis of a fetal pup at with green outlines indicating tracking of the epicardium and endocardium borders through (E) end diastole and (F) end systole. R: right; L: left.

Speckle-tracking-based stress strain assessments were performed during systole in fetal and adult offspring groups of all dams (Table 2 and Additional file 1: Tables S6,7). Long-axis measurements in adult offspring demonstrated that endocardial global longitudinal strain (GLS) was significantly lower in PPME WT offspring whose dams were exposed to nano-TiO₂ compared to sham-exposed (Table 2). The PPME WT pups also had a significantly lower GLS than PPME Tg adults, who were both progeny of nano-TiO₂ exposed dams. (Table 2). A group-dependent effect was seen with MPME progeny having higher radial and longitudinal strain rates, and diminished longitudinal displacement compared to PPME progeny. Group-dependent effects were also observed in the parasternal short-axis between the MPME (Additional file 1: Table S6). Figures 3(C) and D provide representative B-mode images of a fetal pup in utero during end diastole and end-systole, respectively, as well as speckle-tracking traces correlating to each of these frames (Figure 3(E,F)). At the fetal stage, short-axis radial and circumferential strain rates were significantly lower in both the PPME WT and Tg groups whose dams were exposed to nano-TiO₂ than the pups of the same respective genotype whose dams were sham-exposed (Additional file 1: Table S7). Further, these data also revealed a group-dependent effect based on maternal genotype (MPME vs. PPME). Overall, these changes appear to suggest that enhanced maternal antioxidant protection may provide sufficient protection to thwart cardiac maladaptation that is precipitated by ENM exposure.

Table 2.

Adult progeny systolic stress-strain in the long-axis.

Parameter	Units	PPME WT Shama	PPME WT TiO2b	PPME Tg Shamc	PPME Tg TiO2d	MPME WT Shame	MPME WT TiO2f	MPME Tg Shamg	MPME Tg TiO2h
Radial velocity	Pk cm/s	2.01 ±0.17	1.89 ± 0.12	1.82 ±0.10	2.52 ± 0.16 c	2.03 ± 0.21	2.46 ± 0.23	2.23 ± 0.15	2.08 ± 0.26
Radial displacement	Pk mm	0.47 ± 0.03	0.43 ± 0.03	0.44 ± 0.04	0.61 ± 0.02 a,b,c	0.40 ± 0.00	0.55 ± 0.04	0.47 ± 0.05	0.52 ± 0.03
Radial strain	Pk %	27.82 ± 1.84	25.95 ± 2.85	25.54 ± 3.37	31.55 ± 2.74	26.48 ± 2.58	32.90 ± 2.72	30.69 ± 3.08	23.67 ± 1.49
Radial strainrRate	Pk 1/s	9.10 ±0.83	8.20 ± 0.67	8.62 ± 1.13	11.12 ± 1.16	11.34 ± 1.55	11.80 ± 1.02	11.31 ± 0.34	8.50 ± 0.97
Longitudinal velocity	Pk deg/s	1.14 ± 0.16	1.19 ± 0.18	1.00 ± 0.18	1.07 ± 0.20	1.16 ± 0.36	1.09 ± 0.28	1.28 ± 0.60	0.95 ± 0.15
Longitudinal displacement	Pk deg	0.18 ± 0.01	0.12 ± 0.04	0.15 ± 0.06	0.11 ± 0.02	0.04 ± 0.01	0.06 ± 0.01	0.06 ± 0.01	0.12 ± 0.04
Longitudinal strain	Pk %	−10.99 ± 2.13	−9.51 ± 2.00	−9.16 ± 0.90	−15.81 ± 3.41	−5.38 ± 2.61	−7.98 ± 1.52	−12.99 ± 2.78	−10.68 ± 4.73
Longitudinal strain rate	Pk 1/s	−7.63 ± 0.58	−6.86 ± 0.76	−7.06 ± 1.01	−9.34 ± 0.76	−7.45 ± 0.85	−8.93 ± 1.35	−14.26 ± 1.35 e,f,h	−7.49 ± 1.18
Global longitudinal strain (GLS)	AU	24.26 ± 1.54	16.49 ± 1.97 a,d	19.85 ± 2.20	28.51 ± 1.88 c	20.07 ± 2.25	23.58 ± 3.52	22.67 ± 1.86	22.29 ± 2.44

Peak strain and strain-rate values acquired from B-mode images for longitudinal and radial dimensions in adult progeny (11 weeks of age) following maternal inhalation exposure. Bold text indicates significant data defined by p ≤ 0.05. A two-way ANOVA determined statistical difference between PPME and MPME denoted with underlined values for that parameter. A letter next to a given group denotes statistical significance between the marked group and the group represented by each letter based on a Tukey’s multiple-comparisons test. All data are presented as the mean ± the standard error of the mean (SEM). PPME WT Sham, n = 7; PPME WT TiO₂, n = 6; PPME Tg Sham, n = 6; PPME Tg TiO₂, n = 5; MPME WT Sham, n = 4; MPME WT TiO₂, n = 5; MPME Tg Sham, n = 5; MPME Tg TiO₂, n = 4.

PPME: paternal mPHGPx maternal exposure; MPME: maternal mPHGPx maternal exposure; WT Sham: wild-type offspring whose dam was exposed to control air; Tg Sham: mPHGPx transgenic offspring whose dam was exposed to control air; WT TiO₂: wild-type offspring whose dam was exposed to nano-TiO₂; Tg TiO₂: mPHGPx transgenic offspring whose dam was exposed to nano-TiO₂; GLS: global longitudinal strain; Pk: peak.

ETC complex activities

ETC complex activity assays were performed on protein lysate samples of adult offspring (Figure 4). Two-way ANOVA analysis revealed no significant difference between PPME and MPME offspring for complex I (Figure 4(A)) and complex III activities (Figure 4(B)), but MPME offspring had significantly higher activities for both complex IV (Figure 4(C)) and V (Figure 4(D)). Individual t-tests were implemented to better delineate exposure specific differences between offspring of the same genotype (WT or Tg) and the same cohort (PPME or MPME). The PPME WT offspring of ENM-exposed dams had significantly diminished complex I (Figure 4(A)) and IV (Figure 4(C)) activities than PPME WT offspring whose dams were sham-exposed. These findings are consistent with previously published data from our laboratory (Kunovac et al. 2019).

Figure 4.

Electron transport chain (ETC) complex activities of adult PPME and MPME offspring. (A–D) Electron transport chain (ETC) activities assessed in cardiac protein lysate of adult offspring for (A) complex I, (B) complex III, (C) complex IV, and (D) complex V (ATP Synthase). PPME WT Sham, n = 7; PPME WT TiO₂, n = 6; PPME Tg Sham, n = 6; PPME Tg TiO₂, n = 5; MPME WT Sham, n = 4; MPME WT TiO₂, n = 5; MPME Tg Sham, n = 7; MPME Tg TiO₂, n = 4. Adult = 11 weeks old. Statistical difference was defined by p ≤ 0.05. * = group difference determined by a two-way ANOVA and ns=no statistical difference. A letter above a group denotes statistical significance between those groups based on a Tukey’s multiple-comparisons test. A dagger (†) above a group indicates statistical significance between WT Sham and WT TiO₂ or Tg Sham and Tg TiO₂ of the same maternal genetic group (MPME or PPME) based on a Student’s t-test. All data are presented as the mean ± the standard error of the mean (SEM). Adult = 11 weeks of age, mPHGPx=mitochondrial phospholipid hydroperoxide glutathione peroxidase, PPME = Paternal mPHGPx maternal exposure, MPME = Maternal mPHGPx maternal exposure, WT Sham = wild-type offspring whose dam was exposed to control air, Tg Sham = mPHGPx transgenic offspring whose dam was exposed to control air, WT TiO₂ = wild-type offspring whose dam was exposed to nano-TiO₂, Tg TiO₂ = mPHGPx transgenic offspring whose dam was exposed to nano-TiO₂, Unit = nanomoles of substrate oxidized (minute⁻¹).

GPx activity and H₂O₂ production

Total GPx enzymatic activity (including all isoforms) was measured in adult offspring. MPME offspring had a significantly higher level of total GPx activity than the PPME offspring (Figure 5(A)). Within the PPME group, Tg offspring of sham-exposed dams had elevated enzymatic activity compared to their WT littermates (Figure 5(A)). To assess specific mPHGPx activity, phosphatidylcholine (PCOOH) was used as a substrate. Within the PPME group, the Tg sham adult offspring had the highest level of mPHGPx activity and was significantly higher than the other three PPME groups, including the PPME Tg offspring whose dams were nano-TiO₂ exposed (Figure 5(B)). In the MPME group, the Tg sham and Tg nano-TiO₂ groups were both significantly higher than their littermate controls (Figure 5(B)). Student’s t-tests were used to delineate exposure-specific differences between offspring with the same genotype from the same cohort (PPME or MPME). These analyses revealed that in the PPME group, gestational exposure to ENM resulted in lower mPHGPx activity regardless of the offspring’s genotype (Figure 5(B)).

Figure 5.

Antioxidant activity and H₂O₂ production in adult offspring following gestational inhalation exposure. (A) Total GPx activity in cardiac protein lysate determined by cumene hydroperoxide as substrate and normalized to protein content; (B) Mitochondrial GPx (mPHGPx) activity in cardiac protein lysate determined by phosphatidylcholine hydroperoxide (PCOOH) as substrate and normalized to protein content; (C) Total hydrogen peroxide (H₂O₂) concentration in cardiac protein lysate normalized to protein content; (D) Hydrogen peroxide (H₂O₂) concentration in cardiac isolated mitochondria of adult offspring normalized to protein content. PPME WT Sham, n = 7 (protein) and n = 6 (mitochondria); PPME WT TiO₂, n = 6; PPME Tg Sham, n = 6 (protein) and n = 5 (mitochondria); PPME Tg TiO₂, n = 5; MPME WT Sham, n = 4; MPME WT TiO₂, n = 5; MPME Tg Sham, n = 7; MPME Tg TiO₂, n = 4. Statistical difference was defined by p ≤ 0.05. * = group difference determined by a two-way ANOVA and ns = no statistical difference. A letter above a group denotes statistical significance between those groups based on a Tukey’s multiple-comparisons test. A dagger (†) above a group indicates statistical significance between WT Sham and WT TiO₂ or Tg Sham and Tg TiO₂ of the same maternal genetic group (MPME or PPME) based on a Student’s t-test. All data are presented as the mean ± the standard error of the mean (SEM). Adult = 11 weeks of age, mPHGPx = Mitochondrial phospholipid hydroperoxide glutathione peroxidase, PPME = Paternal mPHGPx maternal exposure, MPME = Maternal mPHGPx maternal exposure, WT Sham = wild-type offspring whose dam was exposed to control air, Tg Sham = mPHGPx transgenic offspring whose dam was exposed to control air, WT TiO₂ = wild-type offspring whose dam was exposed to nano-TiO₂, Tg TiO₂ = mPHGPx transgenic offspring whose dam was exposed to nano-TiO₂.

Previously published data from our laboratory have indicated that maternal ENM exposure during gestation resulted in increased H₂O₂ levels at the fetal stage, which was not sustained into adulthood (Kunovac et al. 2019). In the current study, H₂O₂ levels were unchanged in adult offspring (Figure 5(C)), which was consistent with our previous findings. H₂O₂ levels were also evaluated in isolated mitochondria. H₂O₂ levels were elevated in PPME WT adult offspring whose dams were gestationally exposed to ENM when compared to PPME WT adult offspring whose dams were gestationally-sham-exposed (Figure 5(D)). These data overall indicate that maternal exposure to nano-TiO₂ during gestation results in decreased hydroperoxide scavenging and increased ROS, while maternal antioxidant protection has the potential to prevent these detriments.

Global and mPHGPx M⁶A methylation

Alterations in the relative abundance of m⁶A modifications can affect RNA metabolism and are associated with various diseases, including metabolic diseases affecting the heart (Li et al. 2020). The SEC insertion sequence (SECIS), which is present in the 3′-UTR of mPHGPx and allows for the insertion of the crucial selenocysteine, contains potential m⁶A sites (Xuan et al. 2018, Liu et al. 2020). To investigate whether altered total m⁶A methylation plays a role in the diminished mPHGPx activity in offspring whose dams were exposed to nano-TiO₂, we quantified global m⁶A in RNA isolated from the hearts of adult offspring (Figure 6(A)). Initial analyses indicated that global m⁶A status was statistically unchanged between the groups, but individual t-tests revealed that MPME Tg offspring whose dams were exposed to nano-TiO₂ had significantly lower levels of m⁶A methylation than MPME Tg offspring whose dams were sham-exposed (Figure 6(A)).

Figure 6.

Epitranscriptomic mechanism contributing to diminished antioxidant scavenging ability following maternal nano-TiO₂ inhalation exposure. (A) Global m⁶A content was determined as a percentage (%) of total RNA isolated from adult offspring cardiac tissue. PPME WT Sham, n = 6; PPME WT TiO₂, n = 5; PPME Tg Sham, n = 6; PPME Tg TiO₂, n = 5; MPME WT Sham, n = 4; MPME WT TiO₂, n = 5; MPME Tg Sham, n = 6; MPME Tg TiO₂, n = 4; (B) Isolated RNA with 3 min fragmentation utilized to achieve ~130 nt fragments (top) compared to no fragmentation (bottom) (C) Schematic of the predicted m⁶A site in the 3′-UTR of the mouse mPHGPx mRNA, specifically within the SECIS region; (D) M⁶A-RIP-qPCR for mPHGPx (3′-UTR region including potential m⁶A site). N = 4 per group (ran in triplicate). Statistical difference was defined by P ≤ 0.05 (ns = no significance) based on a one or two-way ANOVA, where appropriate. A letter above a group denotes statistical significance between those groups based on a Tukey’s multiple-comparisons test. A dagger (†) above a group indicates statistical significance between WT Sham and WT TiO₂ or Tg Sham and Tg TiO₂ of the same maternal genetic group (MPME or PPME) based on a Student’s t-test. All data are presented as the mean ± the standard error of the mean (SEM). Adult = 11 weeks old, mPHGPx = mitochondrial phospholipid hydroperoxide glutathione peroxidase, M⁶A = N⁶-Methyladenosine, PPME = Paternal mPHGPx maternal exposure, MPME = Maternal mPHGPx maternal exposure, WT Sham = wild-type offspring whose dam was exposed to control air, Tg Sham = mPHGPx transgenic offspring whose dam was exposed to control air, WT TiO₂ = wild-type offspring whose dam was exposed to nano-TiO₂, Tg TiO₂ = mPHGPx transgenic offspring whose dam was exposed to nano-TiO₂, RIN = RNA integrity number, SECIS = selenocysteine insertion sequence.

In order to evaluate the SEC region of mPHGPx in WT offspring, RNA was fragmented and used to perform an m⁶A RNA immunoprecipitation, followed by qPCR for a specific region of the mPHGPx 3′-UTR. RNA fragmentation incubation time was determined using the Bioanalyzer 2100. A 3-minute incubation was deemed the most appropriate in order to achieve ~130 nt fragments (Figure 6(B)) (Barros-Silva et al. 2020). RMBase v2.0 software predicted the presence of m⁶A sites at the 3′-UTR of mPHGPx mRNA specifically within the SECIS region (Figure 6(C)). RT-qPCR revealed that PPME WT adult offspring whose dams were exposed to nano-TiO₂ had significantly higher m⁶A levels in the 3′-UTR region of mPHGPx than the control group (PPME WT sham) (Figure 6(D)), as well as both the sham and nano-TiO₂ MPME WT groups. These data indicate that gestational exposure in WT dams, has the ability to modulate epitranscriptomic reprogramming that is present in offspring at adulthood. The increase in m⁶A enrichment in the PPME WT offspring of ENM-exposed dams may be responsible for the persistent mitochondrial bioenergetic and cardiac functional detriments. Additionally, enhanced maternal antioxidant capacity may provide protection from the initial stimuli responsible for the altered m⁶A status of mPHGPx.

Discussion

With the inevitable increase in production and application of ENM-enabled products in manufacturing, biomedicine, and general consumer use, a thorough understanding of the longitudinal repercussions is necessary for the development of preventative and protective strategies for workers and consumers. This is of particular importance during fetal growth, where a toxicological insult may elicit immediate and long-term deleterious effects. In this study, we have demonstrated that increased antioxidant defense in the maternal environment is capable of providing preservation to cardiac contractile function potentially through the maintenance of mitochondrial bioenergetics. Furthermore, the presence of the maternal transgene preserved m⁶A status at the 3′-UTR of mPHGPx in offspring despite gestational ENM inhalation exposure. Our data point to a maladaptation in the epitranscriptomic signature as a contributing mechanism responsible for diminished bioenergetics and cardiac contractile deficits in adult offspring following maternal exposure.

Left ventricular ejection fraction (LVEF) has been widely used as a key measure of cardiac contractile function and as a prognostic tool for heart failure (Ponikowski et al. 2016). The dependence of this parameter on load, can limit reproducibility, necessitating a more reliable marker of cardiac function (Kalam, Otahal, and Marwick et al. 2014; Amzulescu et al. 2019). Global longitudinal strain (GLS), a measure of myocardial deformation, is the most commonly utilized clinical application of strain imaging (Ersboll et al. 2013; Mani 2019). GLS appears to be a better predictor of cardiovascular outcomes than LVEF, with better reproducibility (Ashish et al. 2019; Karlsen et al. 2019). Decreases in GLS, as seen in the PPME WT offspring following exposure, highlight the susceptibility of this group to future cardiovascular events (Table 2). The preservation of GLS in the MPME group is therefore noteworthy as the antioxidant protective strategy provided by the mother was able to sufficiently protect her offspring from long-term adverse cardiovascular complications.

Maternal ENM exposure during gestation interferes with fetal development and has long-term consequences for offspring that persist into adulthood (Sun et al. 2013; Hougaard et al. 2015; Bommarito, Martin, and Fry 2017; Crispi, Miranda, and Gratacos 2018). The importance of the maternal environment are outlined by the developmental origins of health and disease (DOHaD) hypothesis and supported by an increasing number of studies (Swanson et al. 2009). The current study highlights how amplifying antioxidant capacity in the maternal environment can deter the maladaptive cardiovascular changes that persist into adulthood, regardless of the progeny’s genotype. Although transgenic expression provides a limitation to clinical uses of maternal antioxidant protection, we establish a basis for investigating other potential antioxidant protective strategies. One example is mitoquinone mesylate (MitoQ), a mitochondrial-targeted antioxidant that has shown promise in improving mitochondrial dysfunction, reducing hydrogen peroxide production, and enhancing antioxidant enzyme activity (Ribeiro Junior et al. 2018; Park et al. 2020). MitoQ also has the ability to prevent the long-term impacts of maternal cigarette smoke exposure on progeny by mitigating renal total ROS in the mothers and offspring at adulthood (Sukjamnong et al. 2018). Future investigations should aim to determine whether supplementation with antioxidant-boosting therapeutics during pregnancy can provide the same protective effect to progeny as maternal mPHGPx overexpression.

As a relatively new area of research in ENM inhalation exposure toxicology, there are a few studies that investigate the detailed epitranscriptomic mechanisms associated with the exposures. However, a recent study demonstrated that in A549 cells, global m⁶A was decreased following particulate matter exposure as a result of altered m⁶A modulator genes (Cayir et al. 2019). Although this study was performed using lung epithelial cells, it is one of the first studies to highlight the concept that exposure to environmental toxicants has the ability to alter m⁶A RNA methylation profiles. M⁶A has previously been established to preferentially occur around stop codons, with high enrichment in 3′-UTRs of human and mouse transcriptomes (Dominissini et al. 2012; Meyer 2019). Further, the 3′-UTR of mPHGPx is a critical region as it is the location of the Sec incorporation machinery, which enables the catalytic activity of this peroxide scavenging protein (Ingold et al. 2018). To ensure proper function of mPHGPx as a peroxide scavenger, proper selenocysteine (Sec) incorporation is required, which involves the recoding of a UGA codon from “Stop” to “Sec” via a complex that is assembled at a specific region of the 3′-UTR known as the SECIS (Shetty and Copeland 2015; Ingold et al. 2018). Recoding occurs through an eEFSec/Sec-tRNASec complex binding to Sec binding protein 2 (SBP2), which is anchored to the SECIS in the 3′-UTR. Alterations to the SECIS region, may affect the efficiency of Sec incorporation and thus alter the catalytic activity and/or expression of mPHGPx by retaining the premature termination codon, potentially resulting in decay (Squires et al. 2007; Shetty and Copeland 2015). Modulation of this critical region may be the mechanism that results in the persistent adverse outcomes of maternal ENM inhalation exposure into adulthood.

Early studies of m⁶A modifications determined that m⁶A primarily occurs in two possible sequences, Gm⁶AC or Am⁶AC (Wei and Moss 1977), but have further been specified as RRACH (R = A/G and H = A/C/U) (Dominissini et al. 2012; Batista 2017; Xuan et al. 2018). Interestingly, RMBase v2.0 software, which was utilized as a way of scanning for RRACH motifs, predicted m⁶A sites at the 3′-UTR of mPHGPx mRNA specifically within the SECIS region (Xuan et al. 2018), as demonstrated in Figure 6(C). The presence of m⁶A sites in this region was also confirmed in a recent study that aimed to assess the methylome of human and mouse tissue (Liu et al. 2020). Papp et al. reported that inefficient selenoprotein synthesis can increase sensitivity to oxidative stress, which is one potential explanation for the alterations seen at the fetal stage following gestational ENM inhalation exposure (Papp et al. 2006). Moreover, increased methylation, shown in Figure 6, may alter Sec incorporation by changing the secondary structure of the SECIS hairpin (Roundtree et al. 2017) and preventing SBP2 from anchoring the incorporation complex to the RNA (Fletcher et al. 2001). This in turn may be leading to inefficient mPHGPx hydroperoxide scavenging (Ingold et al. 2018) at the adult stage. The decrease in mPHGPx protein expression at the fetal stage compared to the maintenance of expression, but the diminution of mPHGPx activity at the adult stage will require further investigation. A possible explanation is that the epitranscriptomic remodeling occurring at the adult stage may differ from the initial reprogramming that occurred in utero following maternal nano-TiO₂ inhalation exposure.

Figure 7 provides an overview of the mechanisms involved in the long-term ramifications of maternal ENM inhalation exposure during gestation on offspring. Overexpression of mPHGPx in the pregnant dam limits H₂O₂ (ROS) production, thereby mitigating changes to mitochondrial bioenergetics and cardiac contractile function in adult progeny. However, without the overexpression of mPHGPx in the maternal environment, ROS production is not controlled following nano-TiO₂ exposure during gestation, resulting in downstream effects on cardiac function and mitochondrial bioenergetics. This occurs, in part, as a result of elevated m⁶A methylation at the SECIS region of mPHGPx, which disrupts Sec incorporation and diminishes mPHGPx catalytic activity. The current study highlights the therapeutic potential of enhanced antioxidant protection in the maternal environment during gestation (Figure 7). Increased hydroperoxide scavenging through this approach can provide protection for offspring into adulthood from the deleterious effects of maternal particulate exposure during gestation, without the need for supplementation by the progeny. Overexpression of mPHGPx in the pregnant dam likely circumvents the initial surge of oxidative stress that is elicited by the xenobiotic exposure, through a mechanism that involves epitranscriptomic reprogramming.

Figure 7.

Physiological and molecular consequences of gestational nano-TiO₂ inhalation exposure in adult offspring. Following gestational exposure, elevated ROS levels in the progeny are concomitant with cardiac functional alterations that can be mitigated by overexpression of maternal mPHGPx. Augmented mitochondrial ROS plays a key role in the persistence of deficits into adulthood (11 weeks of age) characterized by diminished GLS and ETC complex activities. A mechanism that could promote the sustained consequences is initiated by high oxidative stress, which increases m⁶A methylation at the SECIS in the 3′-UTR of mPHGPx. Ultimately, this may tamper with the SECIS binding region for the SBP2 that is vital for the incorporation of selenocysteine and the catalytic activity of mPHGPx. Decreased mPHGPx activity then propagates mitochondrial bioenergetic deficits, limiting overall cardiac performance in the adult offspring. MPHGPx = mitochondrial phospholipid hydroperoxide glutathione peroxidase, ROS = reactive oxygen species, GLS = global longitudinal strain, ETC = electron transport chain, m⁶A = N⁶-methyladenosine, SECIS = Selenocysteine insertion sequence, SBP2 = SECIS binding protein 2.

At the adult stage, maternal ENM exposure during gestation was associated with altered m⁶A methylation of mPHGPx that contributes to diminished enzymatic activity and persistent cardiac contractile changes. These findings highlight a specific targeted modification to the epitranscriptome that may be mechanistically linked to sustained bioenergetic and cardiac dysfunction associated with gestational ENM exposure. Alterations to the epitranscriptome presented in this study provide a new perspective on the role of m⁶A methylation changes in the fields of cardiovascular ENM inhalation exposure toxicology. Our study introduces a protective strategy that can be implemented to safeguard developing progeny against long-term cardiovascular ramifications.