Maternal nano-titanium dioxide inhalation exposure alters placental cyclooxygenase and oxidant balance in a sexually dimorphic manner

Abstract

The placenta plays a critical role in nutrient-waste exchange between the maternal and fetal circulation, and thus impacts fetal growth and development. We have previously shown that nano-titanium dioxide (nano-TiO₂) inhalation exposure during gestation decreased fetal female pup and placenta mass [1], which persists in the following generation [2]. In utero exposed females, once mated, their offspring’s placentas had increased capacity for H₂O₂ production. Generation of oxidants such as hydrogen peroxide (H₂O₂), have been shown to impact cyclooxygenase activity, specifically metabolites such as prostacyclin (PGI₂) or thromboxane (TXA₂). Therefore, we hypothesized that maternal nano-TiO₂ inhalation exposure during gestation results in alterations in placental production of prostacyclin and thromboxane mediated by enhanced H₂O₂ production in a sexually dimorphic manner. Pregnant Sprague-Dawley rats were exposed to nano-TiO₂ aerosols or filtered air (sham--control) from gestational day (GD) 10–19. Dams were euthanized on GD 20, and fetal serum and placental tissue were collected based on fetal sex. Fetal placental zones (junctional zone (JZ) and labyrinth zone (LZ)) were assessed for xanthine oxidoreductase (XOR) activity, H₂O₂, and catalase activity, as well as 6-keto-PGF₁α and TXB₂ levels. Nano-TiO₂ exposed fetal female LZ demonstrated significantly greater XOR activity compared to exposed males. Exposed fetal female LZ also demonstrated significantly diminished catalase activity compared to sham-control females. Exposed fetal female LZ had significantly increased abundance of 6-keto-PGF₁α compared to sham-control females and increased TXB₂ compared to exposed males. In the aggregate these data indicate that maternal nano-TiO₂ inhalation exposure has a greater impact on redox homeostasis and PGI₂/TXA₂ balance in the fetal female LZ. Future studies need to address if treatment with an XO inhibitor during gestation can prevent diminished fetal female growth during maternal nano-TiO₂ inhalation exposure.

Introduction

Historically, the placenta has been considered an asexual organ, however the fertilized egg gives rise to both fetus and placenta, thus the placenta and fetus are the same sex [1,3,4]. Sex differences in fetal development and growth have long been established, with males having accelerated growth rates relative to females, which can be observed as early as the blastocyst stage [5–7]. Additionally, growth of the placental unit is also sex dependent. In humans, male placentas exhibit greater placental growth (mass) compared to females, which facilitates the increased growth demands of the male fetus [8]. Blood flow to the placenta also differs between sexes with male fetuses from normotensive pregnancies having increased blood flow over time compared to female fetuses [9]. Sex also impacts placental nutrient transfer. Female placentas exhibit increased fatty acid oxidation, purine degradation, and uric acid levels, indicating that females may be more susceptible to oxidative stress [10]. Additional disease states [11,12] or environmental insults [13,14] can further alter and exacerbate oxidative stress within the placenta.

Oxidants are produced through aerobic metabolism [15] and can impair cellular function through damage to proteins, lipids, and DNA [16]. To combat oxidants, the body produces antioxidants to maintain a homeostatic balance of these redox compounds [15]. This balance is especially important in early gestation as redox signaling is required for placentation [15]. A study of hexavalent chromium (a by-product of industrial processes involving stainless steel or metals containing chromium [17]) exposure during gestation in humans found that male placentas had increased oxidative stress markers compared to females and exhibited a greater reduction in antioxidant activity of the enzymes, catalase and superoxide dismutase (SOD) [18]. Hydroxyl radicals (HO^• and hydroperoxyl, HO₂^•) can result in lipid peroxidation and HO^• is thought to be formed via the interaction of H₂O₂ and free iron (Fe²⁺) [19]. The hydroperoxyl radical, HO₂^•, can yield H₂O₂ which can further react with metals, such as iron or copper, to further generate more HO^• [19]. To add to this complex system, COX, along with lipoxygenase, have been demonstrated to be minor endogenous sources of these oxidants [20–23]. The rat and human placenta have been shown to express increasing cyclooxygenase (COX)-1 and COX-2, with COX-2 expression towards parturition [24]. In a mouse model, oxidative stress increased close to term and the authors proposed that this played a physiologic role by increasing prostaglandin synthesis, some of which helped to induce parturition [25]. Nano-TiO₂ (a white powder used for its optical properties in paint or photocatalyst [26]) inhalation exposure during gestation increased maternal liver oxidant load and placentas from nano-TiO₂ exposed dams had greater H₂O₂ production capacity [2]. This indicates that nano-TiO₂ exposure during gestation may increase oxidant balance.

Maternal nano-TiO₂ inhalation exposure during gestation increased inflammatory signals from the maternal lung, such as interleukin (IL)-4, -5, and -13 [27]. We also previously demonstrated increased prostacyclin synthase and decreased prostacyclin receptor mRNA expression in the exposed dam lungs [28]. Pulmonary inflammatory signals may enter the circulation due to the interdependence of the pulmonary and cardiovascular systems and therefore act systemically throughout the body. Nano-TiO₂ inhalation exposure decreased endothelium-dependent relaxation in thoracic and femoral arteries [26], uterine radial arterioles [29], and increased uterine radial arteriole vasoconstriction in response to TXA₂ mimetic, U46619 [28]. Additionally, maternal nano-TiO₂ inhalation exposure during gestation increased maternal hepatic mass and H₂O₂ production capacity [2], and decreased GSH: GSSH ratio. Fetal female placental outflow [1] was decreased in response to cyclooxygenase metabolites after maternal nano-TiO₂ inhalation exposure during gestation. Hepatic oxidant production is linked to vascular oxidant production and atherosclerosis [30]. Changes in redox balance during gestation may underlie the hindered fetal development and increased cyclooxygenase vasoreactivity. Whole, placental samples had increased mass [31] and H₂O₂ production capacity [2] from nano-TiO₂ exposed dams (highlighted in Fig. 1). Due to the inflammatory changes within the lungs, microcirculatory adaptations in COX metabolites, and liver oxidant signaling changes, we sought to investigate the changes that occurred within distinct placental zones and how fetal sex impacted these changes. This led to our hypothesis that maternal nano-TiO₂ inhalation during gestation increased placental production of PGI₂ and TXA₂ through H₂O₂ in a sexually dimorphic manner.

Fig. 1. Graphical representation of lung-liver-placenta axis interaction.

Maternal nano-TiO₂ inhalation exposure increases inflammatory signals within the maternal lung, which may spill over into the circulation. Maternal nano-TiO₂ inhalation exposure has been shown to decrease endothelium-dependent relaxation in many vascular beds. Further, maternal hepatic mass and H₂O₂ production capacity is increased in nano-TiO₂ exposed dams. Changes in redox balance during gestation may underlie hindered fetal female development and increased cyclooxygenase vasoreactivity in fetal female placental outflow.

Materials and methods

Nanomaterial

Nano-TiO₂ powder was obtained from Evonik (P25 Aeroxide TiO₂, Parsippany, NJ) and is composed of a mixture of anatase (80 %) and rutile (20 %) TiO₂. Particle characteristics have previously been determined, including primary particle size (21 nm), specific surface area (48.08 m²/g), and Zeta potential (−56.6 mV) [32].

Inhalation exposure and aerosol characterization

A high-pressure acoustical generator (HPAG, IEStechno, Morgantown, WV) created nano-TiO₂ aerosols from bulk TiO₂ dust. These aerosols were fed into a Venturi pump (JS-60M, Vaccon, Medway, MA) to further de-agglomerate the particles. The aerosols were passed into a stainless-steel whole-body exposure chamber (Cube 150, IEStechno, Morgantown, WV) where concentration levels were monitored with a light scattering device (pDR-1500; Thermo Environmental Instruments Inc, Franklin, MA) and regulated with software driven feedback loops to maintain a stable mass aerosol concentration. Real-time aerosol size distributions were regulated in the exposure chamber at a target mass concentration of 12 mg/m³ for 6 h each exposure day. Similar exposure chambers were utilized only for sham-control dams which were exposed to HEPA-filtered air only. In order to increase the likelihood of a viable pregnancy and pups to study, pregnant rats were exposed after implantation during mid to late gestation (gestational day [GD] 10–19) for a total of 6 days. The last exposure occurs on GD 19, 24 h prior tissue collection. The exposure paradigm and aerosol characterization utilized in this study have been previously described [1].

Animal model

Female Sprague-Dawley (SD) rats were purchased from Hilltop Laboratories (Scottdale, PA) and single-housed in an American Association for Accreditation of Laboratory Animal Care (AAALAC) approved facility at West Virginia University (WVU). Temperature (20–26 °C), relative humidity (30–70 %), and light-dark cycle (12:12 h) was maintained for housed rats. After 48–72 h for acclimation, rats were randomly assigned to either sham-control (N = 13) or nano-TiO₂ (N = 14) exposure groups. Rats had ad libitum access to standard chow (2918X; Envigo, Indianapolis, IN) and water throughout the acclimation and exposure periods. Rats were euthanized via thoracotomy and heart removal and then fetal distribution within the uterine horns and fetal sex were recorded. Fetal tissue was weighed and grouped according to fetal sex: placental junctional zone (endocrine production site), and placental labyrinth zone (nutrient-waste exchange), serum, liver, and lung were collected. Placental junctional zone (JZ) and labyrinth zone (LZ) were bluntly dissected [33], flash frozen in liquid nitrogen, and stored at −80 °C until use. All procedures were approved by the WVU Institutional Animal Care and Use Committee.

Briefly, once dams were euthanized and pup count was recorded (previously published [1]), the uterus was surgically excised and placed into a dissection dish containing physiological salt solution (PSS, in mmol/L: 129.8 NaCl, 5.4 KCl, 0.5 NaH₂PO₄, 0.83 MgSO₄, 19.0 NaHCO₃, 1.8 CaCl₂, 5.5 glucose). The uterus was incised longitudinally, and amnionic sacs were opened to allow for quick identification of fetal sex. Fetal sex and position within the horn was recorded, then the first male and female nearest the cervix were removed with the placenta still attached. This unit was utilized for placental hemodynamic assessment and has been previously published [1]. The next set of male and female units were collected for H&E and IHC assessment and has been previously published [1]. The remaining males and females in the litter were utilized for serum and tissue collection for this study.

Serum collection and protein isolation

Fetal blood samples (500 μL) collected by cardiac puncture were collected and pooled based on sex for each litter (BD SST Microtainer Tubes; Becton, Dickinson and Co, Franklin Lakes, NJ). Samples were allowed to clot at 20 °C for 30 min and centrifuged at 10,000 × g for 90 s. Serum was collected and flash frozen until use. JZ and LZ were individually homogenized with a bead mill homogenizer (1.6 mm stainless steel beads; Next Advance, Troy, New York) in phosphate buffered saline (PBS; Fisher Scientific, Waltham, MA) containing a protease and phosphatase inhibitor cocktail (Thermo Fisher Scientific, Waltham, MA). Homogenates were centrifuged at 4 °C and 14,000 × g for 15 min using a tabletop centrifuge (Hettich Lab Instrument AB, Stockholm, Sweden). Supernatant was collected, kept on ice, and total protein concentrations assessed according to the Bradford Method [34]. Isolated protein samples were stored at −80 °C until assayed.

Xanthine oxidoreductase activity

JZ and LZ samples were homogenized in ice cold RIPA containing protease inhibitor cocktail (Sigma) and were briefly spun down. Tissue samples were processed with a sample volume of 10 μL. Potential urate oxidase (uricase) activity was inhibited by the addition of oxonic acid (100 μM) to ensure UA levels were not altered and thus enzyme activity was not underestimated. Total UA production in 60 min at 37 °C in the presence of xanthine (75 μM) served as the basis for quantification of total xanthine oxidoreductase (XOR) activity. To ensure XO dependence on urate formation as well as establish base-line UA levels, allopurinol (100 μM) was used in parallel samples to inhibit XOR. Following incubation for 60 min at 37 °C, protein was precipitated with ice cold acetonitrile. The samples were centrifuged for 12 min at 13,200 × g, at 4 °C. Following centrifugation, the supernatant was removed, placed in borosilicate glass tubes, dried (60 min), and resuspended in isocratic mobile phase (300 μL) and filtered through 0.20 μm nylon membrane filter unit into 11 mm plastic snap top auto-sample vials. The UA content was measured by electrochemical detection (Vanquish UltiMate 3000 ECD-3000RS) coupled to reverse-phase HPLC using a C18 column (150 × 4.6 mm, 3 μm particle size, Luna Phenomenex) and isocratic mobile phase (50 mM sodium dihydrogen phosphate, 4 mM dodecyl-trimethylammonium chloride, 2.5 % methanol, pH 7.0). One unit of activity (U) was defined as 1 μmole/min urate formed at 37 °C and pH 7.4. Results were expressed as a concentration relative to total sample protein content [35]. Inter- and intra-assay variability were less than 10 %.

Hydrogen peroxide activity assay

Placental zone samples were assessed for hydrogen peroxide activity using the Amplex^™ Red Hydrogen Peroxide/Peroxidase Assay Kit (A22188; Invitrogen, Waltham, MA) and performed according to manufacturer recommendations. Inter- and intra-assay variability were less than 10 %.

Catalase activity assay

Placental zone tissue and serum samples were assayed for catalase activity using the Catalase Activity Assay Kit (ab83464; Abcam, Cambridge, U.K.) and performed according to manufacturer recommendations. Inter- and intra-assay variability were less than 10 %.

Enzyme-linked immunosorbent assays (ELISA)

Fetal tissue extracts and serum were diluted with assay buffer and assays were performed to assess TXB₂ (501020; Thromboxane B2 Elisa Kit) and 6-keto-PGF₁α (515211, 6-keto Prostaglandin F₁α Elisa Kit) according to manufacturer recommendations (Cayman Chemical Company, Ann Arbor, MI). Inter- and intra-assay variability were less than 10 %.

mRNA analysis by real-time PCR

Total RNA was extracted from placental zones, fetal lungs, and resorption sites via the RNeasy Kit based on manufacturers recommendations (74104; Qiagen, Hilden, Germany). Total RNA was then transcribed to cDNA via the High-Capacity RNA-to-cDNA kit (4387406; Thermo Fisher Scientific, Waltham, MA).

Using real-time PCR, prostacyclin synthase (Ptgis), prostacyclin receptor (Ptgir), thromboxane receptor (Tbxa2r), superoxide dismutase (Sod1), catalase (Cat), estrogen receptor alpha (Esr1), and estrogen receptor beta (Esr2) were assessed. Primers were purchased from Integrated DNA Technologies (Coralville, IA); GenBank accession numbers, primer sequences, and references are listed in Table 1. Samples were analyzed in triplicate using iTAQ Universal SYBR Green Supermix (Bio-Rad Laboratories, Hercules, CA). Relative fold changes in expression of candidate genes were obtained using the 2−ΔΔCt method [36]. The Ct values were used to calculate ΔCt values for genes of interest [Ct(test) – Ct(housekeeping)]. A housekeeping panel consisting of beta actin, ribosomal small subunit 18, and GAPDH was run on each tissue. Across tested tissue types, GAPDH was the most stable housekeeping gene and therefore tissues were normalized to GAPDH (gapdh). Statistical representation for each gene is based on fold change with respect to GAPDH.

Table 1.

Real-time PCR primers, their sequence, length, and references.

	Forward primer (5’–3’)	Reverse primer (5’–3’)	Length (bp)	Ref.
Gapdh NM_017008	AGGGCTGCCTTCTCTTGTGAC	TGGGTAGAATCATACTGGAACATGTAG	101	[52]
Esr1 NM_012689	GCACATTCCTTCCTTCCGTC	CTCGTTCCCTTGGATCTGGT	216	[53]
Esr2 NM_012754	TTCCCGGCAGCACCAGTAACC	TCCCTCTTTGCGTTTGGACTA	262	[54]
Cat NM_012520	ACAACTCCCAGAAGCCTAAGAATG	GCTTTTCCCTTGGCAGCTATG	76	[52]
Sod2 NM_017051	CGAGCATGGGTTCCATGTC	CTGGACCGCCATGTTTCTTAG	100	[52]
Tbxa2r NM_017054.1	ATCTCCCATCTTGCCATAGTCC	CCGATGATCCTTGAGCCTAAAG	1880	[55]
Ptgis NM_031557	TGGTGTGGGATCTGCGTACA	CCTCCACTCCATACAGGGTCA	567	[56]
Ptgir NM_001077644	TGCTGGAACATCACCTACGT	GTTTCGAGCATAGGCCACAA	422	[56]

Primers were used in placental junctional zone, labyrinth zone, fetal liver, fetal lung, implantation site, and resorption site tissue.

Statistics

Values for xanthine oxidase, catalase, hydrogen peroxide, results from tissue ELISA’s, and mRNA analyses were assessed via a two-way mixed-effects ANOVA. The two-way mixed-effects ANOVA was utilized to compare exposures and fetal sex in a two-by-two comparison. Litter size was nested within the two-way ANOVA to account for differences in pup number between dams. Exposure effects, sex effects, and an interaction of exposure and sex were assessed. If exposure, sex, or an interaction occurred, then individual group differences could be assessed with post-hoc tests. If statistical significance was noted, then Šídák_’s post-hoc test was used. All statistical analysis were performed with Graph Pad Prism 9 (San Diego, CA). Significance was set at p ≤ 0.05, with N representing the number of litters per group. All data is reported as the mean ± SEM, unless otherwise stated.

Results

Nano-TiO₂ exposed fetal females have increased heart mass and decreased liver mass compared to nano-TiO₂ exposed fetal males

An exposure effect was observed with fetal heart mass, with nano-TiO₂ exposed hearts being significantly heavier (Fig. 2A). A sex effect was seen in heart mass, of which fetal females had significantly heavier hearts. Additionally, there was an interaction between exposure and sex, whereby nano-TiO₂ fetal females had significantly increased heart mass compared to nano-TiO₂ fetal males and sham-control fetal females (Fig. 2A). This same pattern persists with normalized fetal heart mass normalized to total fetal mass. There was an exposure effect, sex effect, and an interaction of exposure and sex. Nano-TiO₂ fetal female normalized heart mass was significantly increased compared to nano-TiO₂ fetal males and sham-control fetal females (Fig. 2B).

Fig. 2. Fetal tissue mass and normalized fetal tissue mass.

Fetal wet tissue mass was weighed and then normalized to fetal weight mass. (A) Fetal heart mass. (B) Normalized fetal heart to fetal mass. (C) Fetal lung mass. (D) Normalized fetal lung to fetal mass. (E) Fetal liver mass. (F) Normalized fetal liver. Sham-control male (N = 10), sham-control female (N = 10), nano-TiO₂ male (N = 10), and nano-TiO₂ female (N = 10). *, p ≤ 0.05 exposure effect. #, p ≤ 0.05 for sex effect.

No exposure effect, sex effect, or an interaction of sex and exposure was seen with fetal lung tissue (Fig. 2C). Normalized fetal lung mass to fetal mass had an exposure effect, which nano-TiO₂ had decreased normalized lung tissue compared to sham-control (Fig. 2D). There was no effect of sex or an interaction of exposure and sex.

Fetal liver mass did not have an exposure effect or sex effect but did have a trend (p = 0.054) towards an interaction of exposure and sex. Nano-TiO₂ fetal female livers trended (p = 0.051) towards significantly decreased in mass compared to nano-TiO₂ exposed fetal male liver mass (Fig. 2E). Normalized fetal liver mass to total fetal mass did not have an exposure effect or sex effect but trended (p = 0.054) towards an interaction of exposure and sex (Fig. 2F). Nano-TiO₂ fetal female normalized liver mass was significantly decreased compared to sham-control fetal female and nano-TiO₂ fetal male normalized liver masses.

XOR activity is decreased in nano-TiO₂ fetal female JZ versus sham-control female JZ

In the JZ, there was a significant effect of exposure on XOR activity, with decreased XOR activity in nano-TiO₂ JZ compared to sham-control JZ (Fig. 3A). There was not an effect of fetal sex, nor an interaction of exposure and fetal sex. In the LZ, there was not an effect of exposure on XOR activity, but there was a significant effect of fetal sex on XOR activity. Female LZ XOR activity was significantly increased compared to males. LZ XOR activity was increased in sham-control females compared to sham-control males and in exposed fetal females compared to exposed fetal males (Fig. 3B). There was no interaction of exposure and fetal sex.

Fig. 3. Placental xanthine oxidoreductase activity and hydrogen peroxide concentration.

Placental activity of xanthine oxidoreductase (XOR) and hydrogen peroxide (H₂O₂) concentration were assessed in each placenta zone. (A) Junctional zone (JZ) XOR activity. (B) Labyrinth zone (LZ) XOR activity. (C) JZ H₂O₂ levels. (D) LZ H₂O₂ concentration. Sham-control male (N = 8), sham-control female (N = 6–8), nano-TiO₂ male (N = 8), and nano-TiO₂ female (N = 8). *, p ≤ 0.05 exposure effect. #, p ≤ 0.05 for sex effect.

Hydrogen peroxide levels are decreased in nano-TiO₂ fetal male JZ

In the JZ, there was a trend (p = 0.06) for an exposure effect, where nano-TiO₂ JZ’s exhibited decreased H₂O₂ concentration compared to sham-control JZ’s (Fig. 3C). There was no difference for sex effect or an interaction of exposure and sex. In the LZ, there were no differences among exposure, sex, or an interaction between groups for H₂O₂ concentration, Fig. 3D.

Catalase activity was decreased in nano-TiO₂ fetal female LZ versus sham-control fetal female

An exposure effect was not seen in fetal serum catalase activity. Catalase activity in fetal serum differed in a sex-dependent manner, which levels were greater in males than females. There was no interaction between exposure or sex. Fetal serum catalase activity was significantly less in sham-control females compared to sham-control males (Fig. 4A) and in nano-TiO₂ exposed pups.

Fig. 4. Placental and fetal serum catalase activity.

Catalase activity was assessed in fetal serum and in each placenta zone. (A). Fetal serum catalase activity. (B) JZ catalase activity. (C) LZ catalase activity. Sham-control male (N = 3–4), sham-control female (N = 3–4), nano-TiO₂ male (N = 3–4), and nano-TiO₂ female (N = 3–4). *, p ≤ 0.05 exposure effect. #, p ≤ 0.05 for sex effect.

In the JZ, exposed groups demonstrated significantly decreased levels of catalase activity than nonexposed groups (Fig. 4B), but there was no effect of sex or an interaction of sex and exposure. Catalase activity within the LZ was impacted by exposure as catalase activity was significantly diminished in exposed than sham-control groups. There was also a trend (p = 0.08) toward an effect of sex. Females had increased catalase activity compared to males. LZ catalase activity was decreased in nano-TiO₂ exposed fetal females compared to sham-control fetal females (Fig. 4C). Nano-TiO₂ exposed fetal males catalase activity was not different compared to sham-control males.

Fetal female LZ TXB₂ levels were increased compared to sham-control females

There were significant effects of exposure, sex, and an exposure by sex interaction for circulating levels of fetal 6-keto-PGF₁α. Nano-TiO₂ fetal serum had increased 6-keto-PGF₁α compared to sham-control fetal serum. Fetal females had increased 6-keto-PGF₁α compared to males. Due to an exposure by sex interaction, there was significantly increased levels of 6-keto-PGF₁α in exposed fetal females compared to exposed males and sham-control females (Fig. 5A). In contrast, no effect of exposure was observed in circulating levels of fetal TXB₂. However, TXB₂ demonstrated an effect of sex as levels were increased in fetal females compared to males. TXB₂ was increased for sham-control females compared to sham-control males (Fig. 5B). There was no interaction between exposure and sex.

Fig. 5. Placental and fetal serum 6-keto-PGF_1α and TXB₂ production.

Cyclooxygenase metabolites (6-keto-PGF_1α and TXB₂) were assessed in fetal serum and each placenta zone. (A) Fetal serum levels of 6-keto-PGF_1α. (B) Fetal serum TXB₂ levels. (C) JZ 6-keto-PGF_1α levels. (D) JZ production of TXB₂. (E) LZ levels of 6-keto-PGF_1α. (F) LZ TXB₂ levels. Sham-control males (N = 5), sham-control females (N = 5), nano-TiO₂ males (N = 4–5), and nano-TiO₂ female (N = 3–5). *, p ≤ 0.05 exposure effect. #, p ≤ 0.05 for sex effect.

Levels of 6-keto-PGF₁α in the JZ did not show an effect of exposure, sex, or interaction (Fig. 5C). However, concentrations of TXB₂ in JZ were not impacted by exposure but were significantly impacted by sex. TXB₂ JZ concentration was increased in females compared to males. JZ TXB₂ concentrations had a significant interaction of exposure and sex. JZ TXB₂ levels were significantly increased in exposed fetal females compared to exposed fetal males (Fig. 5D) and sham-control fetal females. There was not a difference between sham-control females and males.

There was a significant effect of exposure on levels of 6-keto-PGF₁α in LZs as values were increased in exposed groups. There was no effect of sex or an interaction of exposure by sex. LZ 6-keto-PGF₁α was increased in exposed fetal females compared to sham-control fetal females (Fig. 5E), the same pattern exists for fetal males. Conversely, LZ TXB₂ levels were significantly affected by exposure, sex, and their interaction. TXB₂ concentrations within the LZ were significantly increased in exposed LZs compared to sham-control. Fetal female LZ TXB₂ levels were increased compared to fetal male LZ’s. TXB₂ concentrations was significantly increased in the exposed fetal female LZ compared to exposed males and sham-control female LZ’s (Fig. 5F).

Exposed fetal female JZ had decreased catalase mRNA expression

JZ prostacyclin receptor gene expression was not impacted by exposure but was significantly affected by sex and the interaction of sex and exposure. Nano-TiO₂ exposed fetal JZ had significantly increased prostacyclin receptor mRNA expression compared to exposed fetal male JZs (Table 2) and sham-control females. JZ catalase expression was not different for exposure, but was significantly impacted by sex and the interaction of sex and exposure. Wherein catalase expression in exposed females was decreased compared to exposed males (Table 2). JZ mRNA expression was not different for thromboxane receptor, prostacyclin synthase, superoxide dismutase, ERα, or ERβ (Table 2). Statistical differences were not detected in LZ mRNA expression for any of the evaluated targets across groups.

Table 2.

mRNA Expression profiles for prostacyclin synthase (PGIS) and receptor (PGIR), thromboxane receptor (TXR), superoxide dismutase (SOD), catalase (CAT), and estrogen receptors (ERα or ERβ) or in placental tissue, JZ and LZ.

Tissue	Group	N	Prostacyclin Synthase	Prostacyclin Receptor	Thromboxane Receptor	Superoxide Dismutase	Catalase	ERα	ERβ
Junctional zone	Sham-control Male	3–5	1.00 ± 0.19	1.00 ± 0.32	1.00 ± 0.21	1.00 ± 0.17	1.00 ± 0.18	1.00 ± 0.73	1.00 ± 0.96
	Nano-TiO2 Male	3–5	0.68 ± 0.22	0.80 ± 0.18	1.13 ± 0.31	0.89 ± 0.09	1.46 ± 0.12	0.59 ± 0.20	1.36 ± 1.10
	Sham-control Female	3–5	1.00 ± 0.33	1.00 ± 0.26	1.00 ± 0.29	1.00 ± 0.09	1.00 ± 0.21	1.00 ± 0.54	1.00 ± 0.15
	Nano-TiO2 Female	3–5	0.58 ± 0.12	2.32 ± 0.37 ab	1.54 ± 0.23	0.85 ± 0.03	0.79 ± 0.09 b	1.21 ± 0.42	3.07 ± 0.08
Labyrinth zone	Sham-control Male	3	1.00 ± 1.40	1.00 ± 0.42	1.00 ± 0.83	1.00 ± 0.27	1.00 ± 0.21	1.00 ± 0.62	1.00 ± 8.56
	Nano-TiO2 Male	3	0.64 ± 0.72	0.76 ± 0.18	1.46 ± 0.31	1.03 ± 0.27	1.16 ± 0.30	1.55 ± 1.02	0.22 ± 1.34
	Sham-control Female	3	1.00 ± 0.28	1.00 ± 0.08	1.00 ± 0.18	1.00 ± 0.13	1.00 ± 0.17	1.00 ± 0.29	1.00 ± 0.18
	Nano-TiO2 Female	3	2.92 ± 1.74	1.27 ± 0.22	0.89 ± 0.04	1.55 ± 0.05	1.52 ± 0.19	1.55 ± 0.88	2.38 ± 1.34

N represents the number of litters per group. Data are fold-change and represent mean ± SEM. JZ TXR, PGIS, SOD, ERα, and ERβ do not have significance for exposure, sex, or an interaction effect (p > 0.05). JZ PGIR has a trend to an exposure effect (p = 0.074), a sex effect (p = 0.019), and an interaction (p = 0.019). JZ CAT does not have an exposure effect (p > 0.05), a trend for a sex effect (p = 0.053), and an interaction (p = 0.053). LZ SOD, TXR, PGIS, PGIR, CAT, ERα, or ERβ did not have an exposure, sex, or interaction effect (p > 0.05).

^ap ≤ 0.05 vs sham-control female.

^bp ≤ 0.05 vs nano-TiO₂ male.

Fetal lung and implantation sites, where the embryo embeds into the uterus and is the last maternal interface that attaches to the placenta, were assessed for cyclooxygenase metabolites, prostacyclin synthase, prostacyclin receptor, and thromboxane receptor. Fetal lung thromboxane receptor mRNA expression expressed a trend for an effect of exposure (p = 0.06). Expression of thromboxane receptor mRNA was decreased in exposed fetal lung compared to sham-control, especially in male nano-TiO2 fetal lung (Table 3). There were no differences in fetal lung mRNA expression of prostacyclin receptor or prostacyclin synthase. Implantation site expression of prostacyclin synthase was not different for exposure but was impacted by sex and the interaction of sex and exposure. Nano-TiO2 exposed fetal female implantation sites had significantly increased prostacyclin synthase mRNA expression compared to exposed male counterparts (Table 3) and sham-control females. There were no significant effects of exposure or sex for implantation sites in mRNA expression of thromboxane receptor or prostacyclin receptor.

Table 3.

mRNA Expression for prostacyclin synthase (PGIS) and receptor (PGIR) and thromboxane receptor (TXR).

Tissue	Exposure/Group	N	Prostacyclin Synthase	Prostacyclin Receptor	Thromboxane Receptor
Lung	Sham-control Male	5	1.00 ± 0.08	1.00 ± 0.23	1.00 ± 0.11
	Nano-TiO2 Male	5	0.94 ± 0.33	0.73 ± 0.21	0.42 ± 0.09
	Sham-control Female	5	1.00 ± 0.26	1.00 ± 0.22	1.00 ± 0.29
	Nano-TiO2 Female	5	0.99 ± 0.07	1.16 ± 0.26	0.87 ± 0.15
Implantation site	Sham-control Male	3	1.00 ± 0.15	1.00 ± 0.18	1.00 ± 0.17
	Nano-TiO2 Male	3	1.85 ± 0.44	1.05 ± 0.35	1.25 ± 0.31
	Sham-control Female	3	1.00 ± 0.92	1.00 ± 0.05	1.00 ± 0.16
	Nano-TiO2 Female	3	5.78 ± 1.27 ab	1.23 ± 0.40	1.48 ± 1.06

N represents the number of litters per group. Data are fold-change and represent mean ± SEM. Fetal lung PGIR and PGIS did not have significance for exposure, sex, or an interaction effect (p > 0.05). Fetal lung TXR had a trend toward significance (p = 0.060) for exposure effect but did not have significance for sex or an interaction effect (p > 0.05). Implantation sites TXR or PGIR did not have an exposure, sex, or interaction effects (p > 0.05). Implantation site PGIS had an exposure effect (p = 0.009), a sex effect (p = 0.043), and an interaction (p = 0.043) effect.

^ap ≤ 0.05 vs sham-control female.

^bp ≤ 0.05 vs nano-TiO₂ male.

^cp ≤ 0.05 vs sham-control male.

Discussion

This study aimed to identify the relationship between H₂O₂ and the cyclooxygenase metabolites, PGI2 and TXA2, in male and female feto-placental units after maternal nano-TiO₂ inhalation exposure during gestation. We have previously demonstrated that maternal nano-TiO₂ inhalation exposure during gestation reduces fetal female mass, female placenta mass and area, as well as placenta immune cell invasion [1]. Liver and placental tissues from dams directly exposed to nano-TiO₂ exhibit increased indices of oxidant stress and an increased H₂O₂ abundance, indicating that these tissues have increased oxidant load [2]. As adults, females exposed to nano-TiO₂ in utero exhibited decreased circulating estrogen levels during gestation, decreased litter size, and decreased pup mass [2]. This suggests that maternal nano-TiO₂ inhalation exposure during gestation not only impacts redox balance in the dam but also impacts female feto-placental growth and development during gestation and later in life. Modifications in redox balance could potentially contribute to the reduced cyclooxygenase metabolite vasoreactivity in dam uterine radial arterioles [28] and/or decreased placental outflow in the presence of U46619, the stable TXA₂ mimetic, in exposed fetal females [1] previously shown. In this study, placentas from female pups whose mothers were exposed to nano-TiO₂ during gestation had increased XOR activity and decreased catalase activity, which indicates an oxidative stress/antioxidant imbalance. Taken together, placentas from female pups exposed in utero demonstrate redox and cyclooxygenase imbalance, impacts placental nutrient-waste exchange, and ultimately fetal growth and development.

It is important to note that toxicant exposure consistently increases oxidant production [37], decreases catalase activity [38], and increases prostacyclin metabolite production [39], in HUVECs or guinea pig alveolar cells. This implies that toxicants, such as nano-TiO₂, can modify the balance between oxidants and antioxidants, thus disrupting downstream signaling by molecules such as prostacyclin or thromboxane. Additionally, placental growth occurs in a hypoxic environment, with oxygen tension residing at 2–3 % (25–38 μM dissolved O₂) in the first trimester and beginning of the second [40]. This becomes important when considering that the low O₂ tension diminishes superoxide generation by xanthine oxidase (XO), thus elevates H₂O₂ production to nearly 90 % [41]. Our current study found that nano-TiO₂ exposed females had increased XDH activity. This will lead to increased oxidants and uric acid production, which may increase placental inflammation and insufficiency, both of which we have previously demonstrated in exposed fetal females [1]. This implies that increased XOR in these exposed fetal females could contribute to their increased placental resistance and modified placental structure, thus impacting fetal female growth and development.

XOR can generate superoxide, H₂O₂, and NADH when transferring electrons from xanthine to oxygen and NAD⁺ [42]. Due to the increased XOR activity found in exposed fetal female placentas, we examined H₂O₂ levels. No differences were observed in H₂O₂ levels across fetal sex or placental zones. This differs with previous results from our laboratory, which demonstrated increased placental capacity to produce H₂O₂ in collected tissue from nano-TiO₂ exposed dams [2]. In a mouse model of gestational exposure to hypoxia, placentas from hypoxic male fetuses were heavier than the hypoxia exposed fetal female placentas [43], and the hypoxic fetal female placentas demonstrated increased oxidative stress markers. This study highlights that oxidative stress is impacted by fetal sex. In homeostasis, redox balance is maintained via the action of small molecule antioxidants and enzymes such as catalase. Herein, we demonstrated that catalase activity in nano-TiO₂ females was significantly decreased within the LZ. Therefore, the decreased growth in female fetuses previously shown to occur due to maternal nano-TiO₂ inhalation exposure [1] may be explained, at least in part, by oxidative imbalance within the LZ, where nutrient-waste exchange occurs.

Previous results from our laboratory demonstrated significantly decreased circulating TXB₂ levels in nano-TiO₂ exposed dams [28]. Based on these results, both 6-keto-PGF₁α and TXB₂ are important vasoactive compounds, thus it was necessary to measure placental zone and fetal serum levels of 6-keto-PGF₁α and TXB₂. Increased circulating levels of 6-keto-PGF₁α were found in the fetal females from nano-TiO₂ exposed dams, which could be indicative of an elevated immune response. Complications during gestation, such as asthma and gestational diabetes, have been associated with increased placental expression of inflammatory immune markers such as, TNF-α, IL-1β, IL-6, and IL-8 [44,45]. Further, Hofbauer cells, which are fetal derived macrophages, have increased production of pro-inflammatory cytokines in gestational diabetes [45]. Maternal nano-TiO₂ exposure significantly increased female placenta Hofbauer cell invasion [1] and maternal circulating IL-1β and TNFα [27]. IL-1β is reported to induce COX-2 expression [46–48], which is the enzyme that catalyzes the production of PGI₂ [48], and could promote this increased 6-keto-PGF₁α levels in fetal females from nano-TiO₂ exposed dams. Placental tissue, specifically the LZ, exhibited significantly increased 6-keto-PGF₁α and TXB₂ levels from fetal females that were exposed in utero. Human pregnancies with intervillous blood flow below the normal mean (less than 130 mL/min) exhibit increased placental TXA₂ production [49]. In a preeclampsia mouse model that simulated modified blood flow to the fetoplacental unit, preeclampsia mice placentas had elevated TXA₂ synthase and plasma TXA₂, but not prostacyclin levels [50]. This indicates that the balance of TXA₂ and PGI₂ levels is critical in regulating blood flow to the feto-placental unit. A rat intrauterine growth restriction (IUGR) study found that IUGR male umbilical arteries had decreased maximal contraction in response to U46619, the TXA₂ mimetic, compared to IUGR females [51]. This highlights that changes in placental production of TXA₂, and PGI₂ to a lesser extent, impact blood flow and further impact fetal growth and development in a sex-based manner. It is possible that the redox imbalance, specifically the increased XOR-catalyzed peroxide production and decreased catalase activity within the LZ, leads to the overall increase in PGI₂ and TXA₂ levels seen in the LZ of exposed fetal females. This would ultimately result in modified uteroplacental blood flow [1], as we have observed previously, and impair fetal growth and development due to modified redox homeostasis.

However, with all this information, there are a few study limitations that need to be addressed. The first limitation is that we do not know where the XO is coming from. A way to address this concern would be the incorporation of an XO-tissue specific knock-out mouse model. We would need to run a few small studies on wild-type mice first, to verify that the pregnant mice would respond to nano-TiO₂ inhalation exposure during gestation similarly to our current rat model. Further, we do not have any febuxostat data in this study, to verify that XO could be driving these changes seen in H₂O₂ and catalase activity after maternal nano-TiO₂ inhalation exposure during gestation. This is already an on-going study in our laboratory. Another limitation that should be addressed is the difference of H₂O₂ levels seen in this study and a previous one from our laboratory [2]. The differences are most likely caused by a few things. First, herein, the placentas were split into the two main zones, JZ (the endocrine production site) and the LZ (nutrient-waste exchange site), previously the whole placenta was assessed. Second, this study assessed the impacts of fetal sex on placental tissue, which fetal sex does matter in relation to placental and fetal tissue assessment. The sex of the fetus was not accounted for in the previous study. Lastly, herein, we utilized an entirely different assay, and measurement for H₂O₂. In this study, we utilized an Amplex Red H₂O₂ kit, which measured the amount of H₂O₂. Whereas the previous study used a Coumarin boronic acid (CBA) assay, which measures the production rate of H₂O₂. These limitations are important to note and are points that are planned to be addressed in future experiments in our laboratory.

In conclusion, this study sought to determine if maternal nano-TiO₂ inhalation exposure resulted in modified redox homeostasis that altered production of PGI₂/TXA₂ within the placental zones in a sexually dimorphic manner. Our exposure paradigm has provided evidence that maternal nano-TiO₂ inhalation exposure modified redox homeostasis within placental zones and impacted fetal female growth and mass, placental development, and hemodynamics [1]. These modifications likely impact COX metabolite production, ultimately modifying fetal growth in a sexually dimorphic manner, with the greatest growth impacts on exposed females (Fig. 6). These adaptations in the exposed fetal female likely ensures survival throughout gestation but could have lasting negative health impacts on females later in life. Future studies should focus on if treatment with an XO inhibitor, such as febuxostat, during gestation can prevent diminished fetal female growth and development during maternal nano-TiO₂ inhalation exposure. This will help elucidate if dysregulated redox homeostasis underlies the fetal and placental deficits observed after exposure and provide a possible treatment option to women who have imbalanced redox and COX metabolites.

Fig. 6. Summary of the impacts of maternal nano-TiO₂ inhalation exposure during gestation on placental oxidant balance in a sex dependent manner.

Maternal nano-TiO₂ inhalation exposure during gestation results in sexually dimorphic modifications to redox balance and thus impacts cyclooxygenase metabolites in the labyrinth zone at GD 20. Females have demonstrated greater impact in their mass (JZ and LZ), area (JZ) compared to exposed males [1]. Females also have increased XDH activity (LZ), decreased catalase activity (LZ), and increased production of both, 6-keto-PGF_1α and TXB₂ (LZ). These resulting changes are likely responsible for the decreased fetal female weight at GD 20, of which progresses to female deficits into adulthood [2].

Abbreviations:

COX

cyclooxygenase

GD

gestational day

H₂O₂

hydrogen peroxide

JZ

junctional zone

LZ

labyrinth zone

PGI₂

prostacyclin

SOD

superoxide dismutase

TXA₂

thromboxane

XDH

xanthine dehydrogenase

XO

xanthine oxidase

XOR

xanthine oxidoreductase

Data availability

Data will be made available on request.