Increased Placental Anti-Oxidant Response in Asymptomatic and Symptomatic COVID-19 Third-Trimester Pregnancies

Despite Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) -induced Oxidative Stress (OxS) being well documented in different organs, the molecular pathways underlying placental OxS in late-pregnancy women with SARS-CoV-2 infection are poorly understood. Herein, we performed an observational study to determine whether placentae of women testing positive for SARS-CoV-2 during the third trimester of pregnancy showed redox-related alterations involving Catalase (CAT) and Superoxide Dismutase (SOD) antioxidant enzymes as well as placenta morphological anomalies relative to a cohort of healthy pregnant women. Next, we evaluated if placental redox-related alterations and mitochondria pathological changes were correlated with the presence of maternal symptoms. We observed ultrastructural alterations of placental mitochondria accompanied by increased levels of oxidative stress markers Thiobarbituric Acid Reactive Substances (TBARS) and Hypoxia Inducible Factor-1 α (HIF-1α) in SARS-CoV-2 women during the third trimester of pregnancy. Importantly, we found an increase in placental CAT and SOD antioxidant enzymes accompanied by physiological neonatal outcomes. Our findings strongly suggest a placenta-mediated OxS inhibition in response to SARS-CoV-2 infection, thus contrasting the cytotoxic profile caused by Coronavirus Disease 2019 (COVID-19).


Introduction
A cause-effect relationship between abnormal placental response, maternal COVID-19 severity, and neonatal outcome has not been established to date. Inflammatory changes in SARS-CoV-2-positive placentae are commonly reported. The prevalence of chronic villitis of unknown origin is higher in placentae delivered by COVID-19 patients than in those of healthy controls [1]. At present, the extent to which the vertical transmission of SARS-CoV-2 occurs and the timing of such transmission are unclear. Although SARS-CoV-2 intrapartum transmission is possible [2], low incidence rates of in utero transmission [3][4][5][6] and early and late adverse obstetric outcomes [7,8] suggest that the placenta may play a critical role in modulating maternal response to SARS-CoV-2 infection. The main variable is placental permissibility to SARS-CoV-2 entry mechanisms initiated by spike protein (S) attachment an asymptomatic (a; n = 14) and a symptomatic (s; n = 15) group. The control group (n = 12) was composed of women with a normal-term, healthy singleton pregnancy who showed no signs of maternal, placental, or fetal disease and who had participated in a study conducted before 2018. At that time, none of the patients in either group were vaccinated against COVID-19. The historical data from the control group served to rule out SARS-CoV-2 exposure. The most common indications for cesarean delivery in the control group were fetal malpresentation, previous cesarean section, and maternal request. Four full-thickness tissue biopsies (n = 4) from case and control placentae were randomly collected from the intermediate area of the basal plate and snap-frozen immediately after delivery. Next, each single biopsy was processed for both mRNA and protein isolation. Each biopsy was analyzed separately from the others and an average per placenta was obtained and reported. Calcified, necrotic, and seriously damaged areas were excluded from collection. Moreover, placental samples (n = 18, six for each clinical group) were preserved in glutaraldehyde combined with formaldehyde for transmission electron-microscopy analyses. Maternal demographics, obstetric, neonatal outcomes, and COVID-19-related maternal symptoms were recorded.

Transmission Electron Microscopy (TEM)
Placental tissue biopsies were fixed by immersion in 1% formaldehyde + 2% glutaraldehyde in phosphate buffer (PB, 0.2 M, pH 7.4) at 4 • C. After washing in PB, they were post-fixed in osmium ferrocyanide for 1 h at 4 • C, dehydrated in graded acetone, and incubated in acetone/Spurr resin (1:1:30 min; 1:2:30 min) and Spurr resin overnight at room temperature. Finally, samples were embedded in Spurr resin in 0.5 mL Eppendorf tubes (24 h, 70 • C). Ultrathin sections were cut with an ultramicrotome (EM UC6, Leica Microsystems, Wetzlar, Germany), collected on uncoated nickel grids (200 mesh), and counterstained for 30 s with Uranyl Less EM Stain and for 30 s with Lead citrate (Electron Microscopy Sciences, Hatfield, PA, USA). Placental sections were observed with a JEM-1400 Flash transmission electron microscope (JEOL, Tokyo, Japan), and images acquired with a high-sensitivity sCMOS camera. A total of eighteen placentae (six from control pregnancies and twelve from COVID-19-infected pregnant women) were evaluated. The COVID-19 group comprised placentae from six a-COVID-19 women and six s-COVID-19 women.

Lipid Peroxidation Measurement
Since ROS are highly reactive and have a very short half-life, direct detection with accuracy and precision in tissue and body fluids is often unfeasible [32]. Counterwise, peroxyl radicals and hydrogen peroxide are relatively stable molecules (with half-lives of seconds to minutes). Therefore, a promising alternative approach to measure oxidative stress in clinical samples is indirect measurement of ROS by examining the oxidative damage these radicals cause to the cell lipids, proteins, and nucleic acids [33]. For the present study, the degree of placenta lipid peroxidation of the plasma membranes was estimated by measuring the Thiobarbituric Acid-Reactive Substances (TBARS) by means of a TBARS Assay Kit (Cayman chemical, Ann Arbor, MI, USA). Absorbance was measured at 535 nM on an ELISA SR 400 microplate reader and the TBARS values were calculated using a Malondialdehyde (MDA) standard curve, prepared by acid hydrolysis of 1,1,3,3tetramethoxypropane. The values are expressed as MDA µM.

RNA Isolation and Real-Time PCR
Total RNA was isolated from frozen placental biopsies using TRI ® reagent (Sigma-Aldrich, Milan, Italy) according to the manufacturer's instructions and then treated with DNase I to remove genomic DNA contamination. Three micrograms of total RNA was reverse-transcribed using a random-hexamer approach (Fermentas Europe, St. Leon-Rot, Germany) and a RevertAid H Minus First Strand cDNA synthesis kit (Fermentas, Cat. No k1632, Leon-Rot, Germany). qRT-PCR reactions were run on a StepOne™ real-time PCR system instrument (Applied Biosystems, Waltham, MA, USA). Gene expression levels of hypoxia-inducible factors 1 α (HIF-1α), CAT, and SOD1 were determined by real-time PCR using specific TaqMan primers and probes following the manufacturer's protocol (Life Technologies, Carlsbad, CA, USA, Cat. No 4331182). TaqMan primers and probes for ribosomal 18S, HIF-1α, CAT, and SOD1 were purchased from Applied Biosystems as TaqMan gene expression assays. For relative quantification, PCR signals were compared between the groups after normalization using ribosomal 18S RNA expression as an internal reference (Life Technologies, Carlsbad, CA, USA, Cat. No 4333760F). Relative expression and fold change were calculated according to Livak and Schmittgen [34].

Assessment of SOD and CAT Enzymatic Activities
CAT and SOD enzyme activity in the placental biopsies was determined using commercially available kits (Cayman Chemical, Ann Arbor, MI, USA) and following the manufacturer's instructions. Briefly, CAT activity was determined by measuring catalase peroxidative function based on the reaction between CAT and methanol in the presence of optimum concentration of hydrogen peroxide. Formaldehyde was measured spectrophotometrically at 540 nm using 4 amino-3-hydrazino-5-mercapto-1,2,4-triazole. Results are expressed in nmol/min/mL. Total SOD activity was measured by reduction of cytochrome C by superoxide (O 2 •−) radicals monitored spectrophotometrically at 450 nm using the xanthine-xanthine oxidase system. Results are expressed in U/mL.

Statistical Analysis
Data are presented as median ± SEM (standard error of the median) and the Kruskal-Wallis non-parametric test was used since data did not show the same distribution. If a significant difference was found between groups, the Mann-Whitney U-test with Bonferroni's correction was performed. Categorical variables are presented as frequency (percentages); the chi-square test was performed for comparison between groups. Statistical analysis was carried out using SPSS Version 27 statistical software (IBM Corp. Released 2020. IBM SPSS Statistics for Windows, Version 27.0. Armonk, NY, USA: IBM Corp.). Significance was accepted at p < 0.05. Table 1 presents the clinical features of the study population. The case and the control groups were comparable for maternal age and gestational age at delivery, and percentage of nulliparous women, cesarean section, and female and male neonates. A total of 29 (70%) pregnant women tested positive for SARS-CoV-2 infection, of which 15 (51%) were categorized as symptomatic (s-COVID-19) according to previously published criteria [35] and 14 (48%) as asymptomatic (a-COVID-19). The control (CTRL) group was composed of 12 women. As expected, the incidence of overweight was higher among the women testing positive for COVID-19; there was no statistically significant difference between the a-COVID-19 and s-COVID-19 groups. No stillbirths were recorded and no neonatal respiratory support was required within 24 h of birth in the SARS-CoV-2-positive women during pregnancy. There were no differences in placental or birth weight or percentage of female and male neonates between groups. However, a significantly higher percentage of female neonates was reported within the a-COVID-19 and s-COVID-19 groups (71.4 and 60%, respectively; p < 0.0001). In the CTRL group, we reported a higher percentage of male neonates (58.3%; p = 0.001). Finally, an increased percentage of abnormal cardiotocography (CTG) was noted in the a-COVID-19 and s-COVID-19 groups (21.4% and 26.7%, respectively) compared to the CTRL group, even though no significant differences were reported between the a-COVID-19 and s-COVID-19 groups.

Placenta Ultrastructural Morphology in a-COVID-19, s-COVID-19, and CTRL Pregnant Women
We used transmission electron microscopy to assess ultrastructural alterations in the placentae of women with COVID-19. Figure 1A,B illustrate the general appearance of trophoblast cells in the placenta of control (A) and SARS-CoV-2-infected (B) women. In both groups, cells were filled with vacuolar structures of variable size interspersed with other cellular organelles. Mitochondria are shown at higher magnification in panels C-F. While in the placentae from CTRL group the mitochondria had a normal appearance ( Figure 1C), ultrastructural alterations were observed in placentae of both a-COVID-19 and s-COVID-19 patients ( Figure 1D-F). Specifically, these organelles frequently displayed swellings with a reduction in the number of cristae ( Figure 1D). Some of the mitochondria presented severe alterations consisting in matrix rarefaction, cristae loss, and the formation of abnormal membranous structures ( Figure 1E,F). These alterations were invariably observed in all placentae from SARS-CoV-2-infected women but not in control placentae. It is of note that viral particles were not detected unambiguously in these samples.

Assessment of Placental Oxidative Stress Markers
There was a significant increase in TBARS levels as a lipid peroxidation biomarker in the a-COVID-19 (p = 0.018, 1.1-fold increase) and the s-COVID-19 group (p = 0.003, 1.1-fold increase) placentae compared to the CTRL group (Figure 2A), which indicated placental OxS onset after SARS-CoV-2 infection. No significant differences were reported between the a-COVID-19 and s-COVID-19 groups (p > 0.05). Placental OxS in the COVID-19-positive women was detected by hypoxia-inducible factors 1 α (HIF-1 α), a key transcription factor that regulates cellular response to hypoxia and plays a critical role in ROS production and OxS onset. We found significant HIF-1 α overexpression in the placentae of the a-COVID-19 (p = 0.004, 1.95-fold increase) and the s-COVID-19 (p = 0.028, 1.87-fold Increase) groups compared to the CTRL group ( Figure 2B). However, no significant differences were reported between the a-COVID-19 and s-COVID-19 groups (p > 0.05).

Assessment of Placental Antioxidant Defense Markers
Endogenous defense against an abundance of pro-oxidant agents involves the overall action of antioxidant enzymes to detoxify the free radicals and avert tissue damage. SOD1 catalyzes the conversion of the O 2 •− radical to H 2 O 2 , then cytosolic CAT transforms H 2 O 2 to water. We noted significantly higher CAT mRNA levels in the placenta of the a-COVID-19 (p = 0.006, 2.16-fold increase) and s-COVID-19 (p = 0.026, 2.19-fold increase) groups compared to the CTRL group ( Figure 3A). However, no significant differences were reported between the a-COVID-19 and s-COVID-19 groups (p > 0.05). CAT enzymatic activity was significantly increased in the a-COVID-19 (p = 0.04, 1.1-fold increase) and s-COVID-19 (p = 0.013, 1.11-fold increase) groups compared to the CTRL group ( Figure 3C), while no significant differences were reported between the a-COVID-19 and s-COVID-19 groups (p > 0.05). CAT overexpression was accompanied by a significant increase in SOD mRNA levels in the placentae of the a-COVID-19 (p = 0.014, 2.12-fold increase) and s-COVID-19 (p = 0.041, 2.36-fold increase) groups compared to the CTRL group ( Figure 3B). SOD gene overexpression was observed, with a significant increase in SOD enzymatic activity in the a-COVID-19 (p = 0.004, 1.12-fold increase) and s-COVID-19 (p = 0.011, 1.1-fold increase) groups compared to the CTRL group ( Figure 3D). No significant differences in SOD gene expression levels and enzymatic activities were reported between the a-COVID-19 and s-COVID-19 groups (p > 0.05).

Comparisons of Placental Oxidative Stress and Antioxidant Defense Markers in SARS-CoV-2-Infected Women with and without Pregnancy-Related Comorbidities
We compared TBARS, HIF-1α, CAT, and SOD levels between COVID-19-positive women with and without complications during pregnancy and the CTRL group in order to rule out the potential contribution of pregnancy-related comorbidity (gestational hypertension, intrauterine growth restriction, preeclampsia) or abnormal CTG-to-OxS markers and anti-oxidant overexpression. We found a significant increase in the OxS markers TBARS (p = 0.031) and HIF-1 α (p = 0.01) ( Figure 4A,B) and anti-oxidant CAT (gene, p = 0.039; enzymatic activity, p > 0.05) and SOD (gene, p = 0.016; enzymatic activity, p = 0.006) ( Figure 4C-F) in the COVID-19 women compared to the CTRL group. There were no significant differences between the SARS-CoV-2-positive women who went through pregnancy without comorbidities and those with a pregnancy-related comorbidity (p > 0.05, Figure 4).

Discussion
To our best knowledge, this is the first report of alterations of placental mitochondria associated with increased levels of oxidative stress markers TBARS and HIF-1α in women with SARS-CoV-2 infection during the third trimester of pregnancy. Our data confirm the known COVID-19 pro-oxidant and cytotoxic profile of the placenta. Importantly, our results were associated with an increase in placental SOD and CAT anti-oxidant enzymatic activities accompanied by physiological neonatal outcomes providing clues for a compensatory adaptation of the placenta to maintain its physiological abilities and to protect fetal growth. Our observational findings strongly suggest placenta-mediated OxS inhibition in response to SARS-CoV-2 infection, thus contrasting the cytotoxic profile caused by COVID-19.
In line with Gao and colleagues [36], we found no evidence of unfavorable obstetric and neonatal outcomes in women infected by SARS-CoV-2 during the third trimester. The clinical manifestations of SARS-CoV-2 in the pregnant women were similar to those seen in the general population. As reported elsewhere [37], the most common symptom in the symptomatic COVID-19 group was fever. Previous studies [9,[37][38][39] reported that SARS-CoV-2 infection predominantly affects pregnant women over the age of 30 years. Maternal COVID-19 has also been linked with iatrogenic preterm birth due to maternal indications, but the overall rates of spontaneous preterm births are not high and the rates of stillbirths and neonatal deaths do not seem to be any higher than the background rates [40][41][42][43]. This suggests that placental trophoblasts may be less susceptible to SARS-CoV-2 infection during the third trimester of pregnancy because of decreased ACE2 expression [37]. Similar to previous reports [44,45], we recorded no cases of vertical transmission of SARS-CoV-2. The most frequent comorbidities that we recorded were gestational overweight, gestational hypertension, and hematological disorders. These observations are shared by other reports [37,39,46].
The strong perturbation of mitochondria morphology that we observed in COVID-19 placentae was previously described in lung epithelial cells [47], and it is often associated with functional alterations comprising inhibited mitochondrial biogenesis, loss of mitochondrial membrane potential (MMP or ∆Ψm), and inhibition of oxidative phosphorylation [47,48]. In a recent publication, it was suggested that SARS-CoV-2 RNA transcripts and open-reading frames (ORFs) such as ORF 9 localize in mitochondria and regulate mitochondrial function [49]. Hijacked mitochondrial functions constitute a favorable condition to increase the steady-state levels of reactive oxygen species (ROS) production [48,50]. These events are recognized as major COVID-19 pathogenic mechanisms suggestive of an altered OxS regulation triggered by SARS-CoV-2 that we confirmed with placental TBARS and HIF-1α overexpression.
As for other beta-coronaviruses's family members, SARS-CoV-2 replicates in the cytoplasm. Nevertheless, we could not unambiguously identify viral particles in placentae from COVID-19 pregnant women. Drastic cytoplasm vacuolization that we reported in SARS-CoV-2 trophoblast cells was previously described in an in vitro model of human conducting airway epithelium and Madin-Darby bovine kidney cells for SARS-CoV [51,52] leading to the hypothesis that vacuoles, in terms of early and late endosomes, could have a key role in the virus assembly process [53]. Accordingly, it was recently proposed that the downregulation of Rab7 small GTPase protein found in placentae from COVID-19 pregnant women could result in retention of the virus in the early endosomes or trapping within late endosomes and MVB, mediating physiological placental blockade of SARS-CoV-2 in pregnancy and consequently its vertical transmission. This hypothesis would help to explain why the presence of SARS-CoV-2 material in the placenta constitutes a rare event and not the rule [54].
Ultrastructural mitochondria alterations are likely correlated with the elevated oxidative stress documented herein by TBARS and HIF-1α overexpression in COVID-19 placentae. This association was previously described in patients with COVID-19 [17,[55][56][57][58][59]. OxS contributes to SARS-CoV-2 pathogenesis and severity by inducing inflammation, loss of immune function, and by increasing viral replication which may result from activation of the nuclear factor kappa B (NF-κB) pathway [58,60]. Moreover, RNA viruses promote changes in the body's antioxidant defense system and affect enzymes such as SOD and CAT [61,62]. Serum CAT and SOD levels were found to be lower in COVID-19 patients than in controls [63]. In contrast, we demonstrated a statistically significant increase in CAT and SOD enzymatic activity in the placenta from COVID-19 women compared to the controls. A plausible explanation is that the placenta mounts a defense mechanism to fight SARS-CoV-2-induced OxS and to ensure physiological fetal growth and development, as we found in the present cohort.
Furthermore, there is ample evidence that the placenta can counteract adverse conditions (e.g., maternal nutritional challenges, glucocorticoid overexposure, hypoxia) [64][65][66]. In detail, the antioxidant activity by the placenta in response to oxidative stress was demonstrated in preeclamptic pregnancies that reached term delivery by making multiple adaptations in mitochondrial function and related processes that were only minimally observed in preeclamptic pregnancies that delivered pre-term [67].
Our study is limited by having included patients who were infected with the first form of SARS-CoV-2 and by the fact that, today, the spread of COVID-19 vaccines has reached acceptable levels in many countries. The immune and placental response in these women may be different. However, there is a high proportion of women who preferred not to undergo vaccination and there are many developing countries where vaccines are not yet widely available. Our observations are therefore of interest in understanding the placental effects of infection in unvaccinated patients.
Researchers interested in studying the placental effects of SARS-CoV-2 will face the possibility that their data will be changed by the effects of immunization. Our experimental model offers a "clean" image of this confounder: the analysis of our samples does not present the risk of presenting different characteristics due to the effect of immunization or infection with subsequent variants (for example Omicron). For these reasons, we consider our results to be original and difficult to be reproduced in the future.

Conclusions
In conclusion, we observed that SARS-CoV-2 infection during the third trimester of pregnancy induced placental mitochondrial alterations, terminal end products of lipid peroxidation, and an antioxidant adaptation most likely to minimize the detrimental effects of COVID-19-induced OxS on fetal development. Our data suggest that the redox-regulated intracellular pathways triggered by SARS-CoV-2 infection may offer a novel therapeutic target for COVID-19 during pregnancy.

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki. The study was approved by the Institutional Review Board (IRB) of the City of Health and Science of Turin (reference number: 00171/2020).

Informed Consent Statement:
Written informed consent for placental collection and subsequent analysis was obtained.

Data Availability Statement:
The raw data supporting the conclusions of this article will be made available by the authors without undue reservation.

Conflicts of Interest:
The authors declare no conflict of interest.