Placental mitochondrial calcium uniporter modulates offspring susceptibility to metabolic dysfunction

Abstract

Mitochondria are crucial for regulating metabolism, but their role in the placenta and how they may shape offspring metabolism and long-term health remains unclear, despite being commonly associated with pregnancy complications. To investigate this, we used a genetic model with placenta-specific deletion of the mitochondrial calcium uniporter (Pl-MCUKO) and assessed the metabolic trajectory of adult offspring. We found that, at baseline, female placental trophoblasts in wild-type animals exhibited higher respiration rates than males. MCU deletion impaired mitochondrial function specifically in female placentas and was accompanied by distinct changes in the metabolomic profiles of protein and lipid metabolism. Transcriptome analysis revealed reduced placental cellular growth pathways, consistent with smaller placentas and reduced embryonic body weights in Pl-MCUKO. Although in utero MCU deletion affected fetal growth, it was insufficient to cause permanent postnatal changes in body weight, as these deficits normalized in adulthood, with normal glucose homeostasis in Pl-MCUKO offspring. However, when challenged with a high-fat diet, Pl-MCUKO females exhibited reduced weight gain, improved glucose and insulin tolerance, smaller fat depots, and increased ambulatory activity compared to controls. This improved metabolic profile was associated with reduced pancreatic β-cell mass but preserved β-cell function. These findings provide direct evidence that placental mitochondrial function can influence the long-term metabolic health of female offspring by modulating key metabolic tissues.

Highlights

• •Deletion of the placental mitochondrial calcium uniporter (MCU) results in reduced placental weight in female offspring, associated with impaired mitochondrial respiration and sex-specific transcriptomic and metabolomic alterations.
• •Female offspring with placental MCU deletion experience fetal growth restriction in utero, followed by a catch-up growth after birth on a normal chow diet.
• •Under high-fat diet, female offspring from MCU-deficient placentas show reduced body weight gain and improved glucose metabolism.

1. Introduction

Pregnancy involves significant physiologic and metabolic changes in the mother's tissues as she adapts to support fetal growth and development. However, these adaptations can be vulnerable to nutrient, hormonal and stress disruptions, leading to pregnancy complications. Maternal conditions such as gestational diabetes mellitus (GDM), preeclampsia, gestational hypertension, and placental insufficiency-induced fetal growth restriction (FGR) pose immediate and long-term risks to both maternal and fetal health. The Barker hypothesis posits that early exposure can influence the health trajectory of offspring. For example, maternal diet can have long-term effects on the metabolic health of the offspring, raising the risk of developing metabolic disorders like obesity and type 2 diabetes later in life [1]. Natural historical events, such as the Dutch famine of 1944, Chinese famine of 1959, and sugar rationing in the United Kingdom during and after World War II, provide striking examples of how prenatal conditions influence long-term health of offspring. Individuals born to mothers who were pregnant during these periods of famine showed a higher risk of cardiometabolic disorders as they aged [2,3]. A recent study showed that individuals exposed to low sugar in the first 1000 days of life have reduced risk for chronic diseases such as diabetes, underscoring the lasting impact of early life exposure on offspring health [4].

At the heart of this complex interplay is the placenta, a critical organ in mediating maternal–fetal exchanges and ensuring optimal fetal development. Serving as a transient organ interface between the mother and fetus, the placenta facilitates nutrient and gas transport, hormone synthesis, and immune modulation, thereby influencing fetal growth and programming [5]. Placental trophoblasts are cells that facilitates nutrient and gas exchange as well as produce key hormones for growth and development of the fetus. Central to the placenta's diverse functions are its mitochondria, dynamic organelles crucial for cellular energy metabolism, redox regulation, steroid hormone synthesis, apoptosis modulation, and various signaling cascades [6]. Placental mitochondria also participate in intricate signaling networks, releasing bioactive molecules like reactive oxygen species (ROS), mitochondrial DNA (mtDNA), and metabolites, which regulate trophoblast proliferation, differentiation, angiogenesis, and immune responses–key processes for placental development and function [[7], [8], [9]].

Placental mitochondria, essential for maintaining optimal health in utero, are often dysregulated in pregnancy complications. For example, preeclampsia is associated with increased mitochondrial content, but these mitochondria show swelling, impaired cristae structure, fragmentation, and heightened ROS production. In cases of FGR, studies report a significant decrease in the expression of key mitochondrial proteins, COXI and COXII, along with reduced mitochondrial content [10,11]. GDM is linked to increased mitochondrial fusion, but these mitochondria exhibit disorganized cristae and elevated free radical levels [[12], [13], [14]]. While mitochondrial disruptions are evident in these pregnancy complications, it remains unclear whether placental mitochondrial dysfunction is a causal factor or consequence of other disrupted processes. To date, no study has directly linked placental mitochondrial dysfunction to offspring metabolic health.

To investigate the role of placental mitochondria in shaping offspring metabolic health, we developed a transgenic mouse model with a placenta-specific deletion of the mitochondrial calcium uniporter (MCU). MCU is a calcium-conducting pore on the mitochondrial inner membrane, functioning as part of a multi-protein complex [15]. The influx of calcium into the mitochondrial matrix, regulated by MCU, is essential for maintaining cellular bioenergetics. This calcium entry activates key enzymes involved in the electron transport chain and TCA cycle, driving crucial energy-producing processes [16,17]. MCU has been reported to play an important role in both normal physiology and disease, particularly in tissues such as the heart, adipose and pancreatic beta-cells, in part through its effects on cellular energetics and redox regulation [[18], [19], [20], [21]]. Single nucleotide polymorphisms (SNPs) in MCU have been linked to altered susceptibility to obesity in humans [22]. In the placenta, MCU expression increases with gestational age and is associated with trophoblast development [8,23]. However, the direct role of MCU in placental function and health remains untested, making it a prime candidate for targeted disruption in placental studies.

Our findings revealed that placental MCU deletion is sufficient to induce mitochondrial dysfunction and FGR, specifically in female offspring. While these offspring experience catch-up growth and reach normal body weight under a standard chow diet, they exhibited reduced body weight gain and improved glucose metabolism when subjected to the metabolic stress of a high-fat diet feeding. This study provides direct in vivo evidence linking placental mitochondrial function to offspring metabolic health trajectory and highlights a sexually dimorphic response of the placenta to mitochondrial disruption.

2. Experimental procedures

2.1. Human placenta procurement & isolated mitochondria seahorse assay from frozen human placenta

A cohort of placenta tissues from appropriate-for-gestational age (AGA) and fetal growth restricted (FGR; defined as an estimated fetal weight and abdominal circumference less than 10th percentile) pregnancies were collected by Dr. Elizabeth Morgan (University of Minnesota-Twin Cities, Baystate Medical Center, University of Connecticut), following published methodology [24,25]. After obtaining informed consent, placenta samples were obtained within 30 min of Cesarean birth, washed briefly in PBS, and snap frozen in liquid nitrogen. All tissue samples were taken in accordance with IRB approval and oversight of Bionet, the tissue procurement arm of the Clinical and Translational Science Institute at the University of Minnesota & Baystate Medical Center. A separate cohort of placenta tissues from women with lean (Healthy), BMI >25 mg/kg (Ob) without underlying maternal disease, and those diagnosed with gestational diabetes requiring insulin (GDM) were collected as part of the University of Minnesota Gestational Origins of Pediatric HEalth Repository (GOPHER): University of Minnesota IRB Approval #STUDY0016978 (Dr. Sarah Wernimont; Principal Investigator). After obtaining informed consent, villous biopsies were obtained within 30 min of Cesarean birth, washed briefly in PBS, and snap frozen in liquid nitrogen. Clinical data was obtained from the electronic medical record. Maternal and fetal characteristics are provided in Supplemental Tables 1–4.

The published protocol for isolating mitochondria and conducting Seahorse analysis from previously frozen tissues was followed as outlined in Osto et al. [26]. Briefly, frozen placentas were homogenized using a glass Dounce homogenizer with MAS buffer. The homogenate was then centrifuged at 1000×g for 10 min at 4 °C, discarding the pellet. After quantifying protein using the BCA assay, 80 μg of the sample was plated in a Seahorse XF96 microplate. The plate was centrifuged at 2000×g for 5 min at 4 °C. Cytochrome C (from equine heart) prepared in MAS buffer was added to each well containing the samples. The Seahorse flux plate was set up with the following injections: A (NADH or Succinate), B (Rotenone/Antimycin A), C (TMPD/Ascorbic acid), and D (Azide). Mitochondrial respiration was measured using the Seahorse XFe96 Extracellular Flux Analyzer (Agilent Technologies, Santa Clara, CA). Mitotracker Deep Red fluorescence dye was used to analyze and normalize Seahorse analysis data.

2.2. Animal model and in vivo mouse procedures

To generate placental specific deletion of target genes, mice with one allele of Cyp19-Cre recombinase (Aromatase Promoter - Cre Recombinase) (from Dr. Gustavo Leone; Medical College of Wisconsin, Milwaukee, Wisconsin, USA) were crossed with following loxP-flanked genes in MCU flox/flox (a generous gift from Dr. John W Elrod; Temple University). LoxP mediated excision leads to frameshift and translation inhibition, leading to abolition of MCU protein [27]. We previously demonstrated that Cyp19-Cre activity is highly specific to the placenta and not in the offspring using a fluorescent Cre reporter. For example, when we deleted mTOR in the placenta, we demonstrated that by western blot analysis that deleted protein of interest (i.e. mTOR) is not altered in key metabolic tissues, including adipose tissue, liver, and pancreatic islets in the offspring [28]. For embryonic tissue collections, SRY genotyping was performed to confirm sex of the fetus. Example of genotyping PCR is presented in Supplemental Fig. 2A. For metabolic studies, offspring were fed 60% kcal high fat diet (HFD; Research Diets, New Brunswick, NJ) for 22–25 weeks, starting at 7-weeks of age.

Glucose (GTT) and insulin tolerance tests (ITT) were conducted in 6-hour fasted animals. Glucose levels were measure immediately preceding and up to 2 h after an intraperitoneal injection of 2 g/kg glucose (Hospira, Pfizer) or 0.5 U/kg insulin (Humalog, Eli Lilly). For in vivo glucose-stimulated insulin secretion (GSIS), serum samples were assayed from facial vein blood collected from overnight fasted mice before and 5 min after a 3 g/kg injection of glucose. Body composition (Echo-MRI-100; QNMR Systems, Houston, TX) and indirect calorimetry including Oxygen consumption (VO₂), carbon dioxide production (VCO₂), Heat, Respiratory Exchange Ratio (RER) and activity (Oxymax Comprehensive Lab Animal Monitoring System (Columbus Instruments, Columbus, OH)) were performed by IBP Physiology Core at University of Minnesota. All animal studies were performed in accordance with the University of Minnesota Institutional Animal Care and Use Committee (IACUC # 2407–42199A).

2.3. Cell culture, trophoblast isolation & seahorse mitochondrial analysis

Bewo cell line (ATCC) were grown and maintained in culture with equal parts F–12K medium (10% FBS, antibiotics) and DMEM medium (10% FBS, antibiotics). Primary murine trophoblasts were isolated following published protocol [29]. Briefly, whole placenta collected at embryonic day 14.5 was digested in buffer containing Collagenase P (0.1%) and DNase (0.002%) at 37C for 1 h. The digested tissues were washed in Medium 199 and filtered through 200 μM nylon filter to remove debris. The cells were then centrifuged through Percoll layer with density gradients 1.028, 1.05, and 1.088 g/ml. Second layer contains the placental trophoblast cells. The cells were plated in 96-well plate and cultured in NTCT-135 medium (10% FBS, antibiotics, 2 mM glutamine, 20 mM HEPES, 10 mM NaHCO). All cells were cultured at 37C in 5% CO2, humidifying incubator.

To test mitochondrial fitness, cells were plated in Seahorse XF96 cell plate. Bewo cells were treated with Benzethonium and Amorolfine (1 uM) for 24 h. Following treatment, media was switched to Seahorse assay medium prepared following the manufacturer's recommendation. Mitochondrial respiration was measured using the Seahorse XF Cell Mito Stress Test Kit for the Seahorse XFe96 Extracellular Flux Analyzer (Agilent Technologies, Santa Clara, CA). Cells were sequentially treated to Oligomycin, FCCP, and Rotenone/Antimycin A. OCR measurements were normalized to DNA measured using the Quant-iT PicoGreen dsDNA Assay.

2.4. Western blot

Protein lysates (40 μg) in RIPA buffer (CST) supplemented with 0.1% SDS and protease/phosphatase inhibitors were quantified using Pierce BCA protein assay. The samples were resolved by SDS-PAGE, transferred to a polyvinylidene difluoride (PVDF) membrane, and blocked with 5% non-fat dry milk. Primary antibodies against MCU (CST), pAMPK T172 (CST), and Vinculin (CST) were applied, followed by IR secondary antibodies. The blot was visualized using LiCor IR detection. Densitometry analysis was performed using NIH ImageJ software.

2.5. Quantitative RT-PCR & RNA sequencing

RNA isolation was performed using the AllPrep DNA/RNA Mini Kit (Qiagen) following the provided instructions (Qiagen). DNA and RNA concentration was measured using a Tecan microplate spectrophotometer. cDNA was synthesized from RNA using a high-capacity cDNA reverse transcription kit (Applied Biosystems). Relative gene expression was evaluated on an Applied Sciences 7900HT Real-Time PCR system using SYBR green (Applied Biosciences) and the ΔΔCT method, with normalization to housekeeping gene. Primer sequences are listed in Table Supplemental Table 5.

RNA samples for sequencing were extracted from e17.5 placenta of male and female control and Pl-MCUKO fetus (n = 4–5) using the RNeasy Plus Micro Kit (Qiagen), following the manufacturer's protocol. RNA integrity (RIN >8) was confirmed using an Agilent 2200 TapeStation. During sequencing, 125 bp paired-end FastQ reads were processed with Trimmomatic (v0.33; Potsdam, Germany), utilizing the optional ‘-q’ parameter and applying a 3 bp sliding window trim from the 3′ end, requiring a minimum quality score of Q30. Quality control of raw sequence data for each sample was conducted using FastQ. Reads were mapped to the UCSC mouse genome (mm10) using Hisat2 (v2.0.2; Dallas, TX, and Baltimore, MD, USA). Gene quantification was performed with FeatureCounts to obtain raw read counts. The Aviti sequencing FASTQ outputs from machine was then inserted into the Minnesota Supercomputing Institute (MSI) Collection of Hierarchical UMII-RIS Pipelines (CHURP, v1.0.1) workflow for bulk RNAseq. Raw matrix counts from the CHURP output was analyzed in R studio (version 4.3.1) with the DESeq2 (version 1.46.0) package. After performing DEseq function to generate the DeSeqDataset object, differential gene expression was determined with results using thresholds of p. adjusted value (FDR) of ≤0.05 with pairwise comparisons. Volcano plots were generated using ggplot2. Heatmaps were generated using pheatmap. For GO and GSEA analysis, the differential gene expression test results table generated from CHURP output was inserted into MSI pathway analysis pipeline run_clusterProfiler.R (version 2.1). GSEA was ranked in Hallmark gene sets of well-characterized biological states or processes.

2.6. Islet isolation & glucose-stimulated insulin secretion

After euthanasia via CO2 overdose and cervical dislocation, the pancreas was perfused with 0.75–1 mg/ml of ice-cold collagenase (Millipore) in Hanks' balanced salt solution through the common bile duct. The pancreas was then inflated, carefully dissected to remove excess fat, and digested for about 8–10 min at 37 °C with manual agitation. The resulting tissue pellet was washed with Hanks' balanced salt solution (Gibco) containing 2% FBS (GenClone), strained to eliminate undigested debris, and filtered through a 70-μm cell strainer (Fisher Scientific). After washing the retained material from the filter, islets were manually isolated and transferred into warm islet media (RPMI supplemented with l-glutamine (Corning), 10% FBS, 100 IU/ml penicillin, 100 μg/ml streptomycin, and 5 mM glucose). These isolated islets were cultured overnight in complete media in a humidified incubator at 37 °C with 5% CO2 before further testing or collection.

For in vitro glucose-stimulated insulin secretion (GSIS), islets were incubated for 2 h in sterile Krebs solution containing (in mM): 114.6 NaCl, 4.7 KCl, 1.2 KH2PO4, 1.1 MgSO4–7H2O, 8 HEPES, 1 CaCl2–2H2O, 10 NaHCO3, and 0.08% BSA w/v. Eight to ten size-matched islets were then placed into 8-μm cell culture inserts and incubated sequentially in low glucose (LG; 2.5 mM) and high glucose (HG; 16.7 mM) solutions for 30 min each. The collected supernatant was analyzed for insulin concentration using ELISA (Alpco). The inserts were also processed to assess insulin and DNA content, normalizing the insulin content of both naïve and post-GSIS islets to DNA.

2.7. Immunohistochemistry & beta-cell mass analysis

5 um thin placenta sections were processed for H&E staining and imaged in brightfield microscope under 4× magnification. 5 um thin pancreas sections (5 representative sections across 200 um apart across pancreas) were incubated in primary insulin antibody, followed by secondary antibodies conjugated to fluorophores and DAPI solution (Thermo Fisher Scientific). Stained slides were imaged on a motorized microscope (ECLIPSE NI-E; Nikon). β-Cell mass was assessed using FIJI software by calculating the ratio of insulin-positive area over total pancreas area (β-cell area) and multiplied by the pancreas weight.

2.8. Untargeted plus metabolomics

Placenta samples (n = 4–5) from littermate control and Pl-MCUKO male and female mice were harvested at embryonic day 17.5. The samples were briefly washed in PBS, snap frozen in liquid nitrogen and submitted to Metware Bio (Woburn, MA, USA) for metabolomic analysis. Samples stored at −80 °C was thawed on ice and homogenized in a ball-mill grinder at 30 Hz for 20s. 400 μL solution (Methanol: Water = 7:3, V/V) containing internal standard was mixed with 20 mg of ground sample and mixed in a shaker at 2500 rpm for 5 min. The mixture was placed on ice for 15 min and centrifuged at 12000 rpm for 10 min (4 °C). 300 μL of the supernatant was collected and placed in −20 °C for 30 min. The sample was then centrifuged at 12000 rpm for 3 min (4 °C). A 200 μL aliquot of the supernatant was used for LC-MS analysis. All samples were for three LC/MS methods. The metabolites were annotated by searching the MetwareBio's in-house database, integrated public database, prediction database and metDNA. Finally, substances with a comprehensive identification score above 0.7 and a CV value of QC samples less than 0.3 were extracted, and then positive and negative mode were combined (substances with the highest qualitative grade and the lowest CV value were retained) to obtain all_sample_data file. From the detected list of metabolites, those classified as exogenous metabolites were excluded for further analysis. For two-group analysis, differential metabolites were determined by VIP (VIP >1) and P-value (P-value <0.05). Metware Bio in-house software was used to analyze and present the data in KEGG metabolic pathway and Venn diagram.

2.9. Statistical analysis

Data are presented as mean ± SEM. Normality was confirmed by the Shapiro–Wilk test, and data were analyzed using 2-tailed nonparametric, unpaired Student t-test. Multiple outcome data were assessed using repeated measures 2-way ANOVA. Statistical analyses were performed in GraphPad Prism version 7 with a significance threshold of p < 0.05.

3. Results

3.1. Placental mitochondrial function is altered in pregnancy complications

Mitochondria are dynamic organelles that play a crucial role in regulating placental energetics. Mitochondrial dysfunction of the placenta is a common observation in various pregnancy complications and placental insufficiency conditions. To assess mitochondrial function in pregnancies with metabolic complications, we conducted Seahorse mitochondrial respiration analysis on isolated mitochondria from previously frozen human donor placental samples from pregnancies from mothers who were lean/healthy, overweight/obese (Ow/Ob), and diagnosed with gestational diabetes mellitus (GDM). Testing the activities of individual electron transport chain (ETC) complexes, in male placentas, no differences in ETC activities were observed across maternal conditions. However, in female placentas, complex I and IV activities were significantly increased in GDM cases compared to both the healthy and Ow/Ob groups (Figure 1A–B; S.Figure 1A–B). Placental mitochondrial mass from GDM mothers was lower than that from healthy and Ob mothers, in male and female placenta respectively (S. Fig 1C). Next, we analyzed a cohort of frozen placental samples from pregnancies classified as either appropriate for gestational age (AGA) or fetal growth restriction (FGR) in both sexes. We observed no differences in the activity of complexes I and IV across the groups. However, complex 2 activity was significantly reduced in female FGR placentas compared to AGA placentas, but not in males (Figure 1C; S.Figure 1D–F). Interestingly, mitochondrial mass in male FGR placenta was lower than that of male AGA placenta (S.Fig 1G). These findings provide an example of altered mitochondrial function that is linked with pregnancy complications, with sex-specific effects. Therefore, studying placental mitochondria and their impact on fetal growth and health is of critical importance.

Figure 1.

Placental mitochondria in disease and gestational age. (A-C) Seahorse mitochondrial respiration analysis of individual electron transport chain complexes (Complex 1, 2, and 4) was conducted on isolated mitochondria from human placentas of male (M) and female (F) births, including those from appropriate for gestational age (AGA) and fetal growth restricted (FGR) infants, as well as from healthy, overweight/obese (Ow/Ob) individuals and those with gestational diabetes (GDM) (n = 4–12). (D-F) Fetal body weight, placenta weight, and placental efficiency (body weight/placenta weight) from male and female mouse fetus at embryonic day 13.5 and 17.5 (average per litter) (n = 20). (G) mRNA expression of mitochondria related genes, normalized to housekeeping gene; presented as Log2FC to M-e13.5 (n = 5). (H–I) Seahorse mitochondrial respiration analysis using mitochondrial stress test kit from Bewo cell line treated with vehicle, 1 uM Benzethonium (Benz), and 1 uM Amorolfine (Amorol) for 24 h, with quantifications (n = 6). Statistical analyses were conducted using 2-way student t-test and 1-way ANOVA with significance ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

3.2. Placental mitochondria are dynamically altered during normal gestation

Placental mitochondria play a vital role in maintaining a normal pregnancy, particularly through supporting the rapid growth and development of the placenta, which is essential for efficient fetal growth [6,30,31]. To explore this, we used C57/Bl6 wild-type mice to investigate changes in fetal and placental weight, as well as the expression of mitochondria-related genes at embryonic days 13.5 (e13.5) and 17.5 (e17.5). Our findings revealed significant increases in fetal body weight from e13.5 to e17.5 for both male and female mice, although female fetuses had a trend toward smaller body weights than males at the e17.5 timepoint (Fig 1D). For placental weight, we observed an increase in male placentas from e13.5 to e17.5, while female placental weight remained unchanged from mid to late gestation (Fig 1E). The placental efficiency, calculated as the ratio of fetal weight to placental weight, was 5–6 times higher at e17.5 compared to e13.5, with no observed sexual dimorphism (Fig 1F). Regarding the expression of mitochondria-related genes, we found no significant differences between male and female placentas (Fig 1G). However, we observed a marked increase over time in critical factors associated with mitochondrial biogenesis, function, and maintenance, including Mcu and its protein complex components, Tfam, Nrf1, Mfn2, and Pparg (Fig 1G). Additionally, we assessed gene expression for placenta cell-type markers and found an increase in endocrine spongiotrophoblasts, glycogen cells, and giant cells in e17.5 placentas compared to e13.5, as indicated by the expression of their respective gene markers, Prl8a8, Gjb3, and Hand1 (S.Fig 1H). We observed no differences in mitochondrial DNA content across gestational ages or between sexes (S.Fig 1I). Overall, these data suggest that mitochondria-related gene expression is upregulated in late gestation, correlating with significant increases in fetal growth and the differentiation of placental cell types.

Mitochondrial dysfunction is a broad term that encompasses a range of impairments, including, but not limited to, reduced ATP synthesis, increased reactive oxygen species production, altered metabolism, and calcium overload. It remains unclear which specific type of mitochondrial dysfunction contributes to placental dysfunction across different pregnancy complications. Previous studies have shown that MCU, the mitochondrial calcium influx regulator, increases throughout gestation [23], and is associated with placental trophoblast differentiation [8]. Rather than targeting a core component of oxidative phosphorylation, such as the ETC, focusing on a regulator of a mitochondrial co-factor such as calcium flux may provide a more physiologically relevant insight into mitochondrial dysfunction in the placenta. To explore this further, we first assessed whether MCU activity could modulate the mitochondrial function of the trophoblast BeWo cell line. We evaluated the mitochondrial function of BeWo cells following treatment with either an MCU activator (Amorolfine) or an inhibitor (Benzethonium) [32]. Our results revealed a significant reduction in both basal and ATP-linked respiration upon MCU inhibition, while activation of MCU led to an increase in respiration (Figure 1H–I; S.Fig 1J). These findings suggest that manipulating MCU effectively alters placental mitochondrial function, establishing it as a potential target for in vivo interventions, which may provide a foundational understanding on the role of mitochondrial dysfunction in placenta-fetal health.

3.3. MCU deletion decreases mitochondrial function in female placenta

To investigate placental mitochondrial dysfunction in vivo and its effects on offspring metabolic health, we generated a mouse model with a placenta trophoblast-specific deletion of MCU, referred here as Pl-MCUKO (Cyp19-cre; MCU flox/flox). Controls (Ctrl) are littermate mice without the Cre genotype. Cyp19Cre is expressed in both the Junctional and Labyrinth zones. Leone et al. generated Cyp19Cre transgenic founders 5912 and 5909 of varying placental specificity [33]. We obtained the 5912 line that showed the highest specificity to the placenta and validated its specificity using GFP reporter [28]. We found GFP was expressed only in the trophoblast lineages of the placenta but not in the fetus and it is not expressed in tissues (islet, liver and adipose) of adult offspring [28]. Our data were later confirmed by Lopez-Tello et al. using a tdTomato reporter [34]. More recently, however, Cyp19-Cre has been reported to exhibit some mosaic expression in the placenta [35]. Therefore, in this study, using a GFP fluorescent Cre reporter, we performed a new experiment in our Cyp19-Cre line and demonstrated high Cre recombination efficiency in both male and female placentas, with the majority of placental cells showing robust GFP expression, consistent with our previous validation [28] (S.Figure 2B). In the placenta, at the protein level, we confirmed a significant reduction in MCU levels from both male and female Pl-MCUKO mice (Figure 2A–B). Lack of complete abolition of MCU via western blot of whole placental tissue may be due to the placenta being composed of multiple non-trophoblast lineage cells, such as immune cells and endothelial cells. Additionally, we assessed MCU expression in key tissues in the offspring (fetal brain, adult ovary and white adipose tissue) and observed no differences in MCU expression between genotypes, suggesting limited off-target Cre action (S.Figure 2C–D). As a proof of concept that MCU deletion impacted mitochondrial function, we assessed the mitochondrial respiration of isolated primary placental trophoblasts ex vivo (Fig 2C). Our analysis revealed sexual dimorphism in placental mitochondrial function in control mice, with female trophoblasts exhibiting higher basal, maximal, and ATP-linked respiration compared to males (Figure 2D–F; S.Figure 2E–G). Despite a comparable reduction in placental MCU protein in our model, we observed a decrease in mitochondrial function specifically in female trophoblasts, with no significant impact on male trophoblasts (Figure 2D–F; S.Figure 2E–G). Additionally, we assessed for AMPK activity, a key master energy sensor in the cell. Here, we observed increased level of the activating phosphorylation site of AMPK (T172) in female P-MCUKO placenta, compared to female control (S.Fig 2H). However, we observed no differences in males between genotypes (S.Fig 2H). These findings suggest that murine female trophoblasts have higher baseline of mitochondrial function than males, and that the deletion of MCU specifically impairs mitochondrial respiration in the female placenta.

Figure 2.

Placenta mitochondrial calcium uniporter deletion leads to mitochondrial dysfunction in females. (A-B) Representative western blot of MCU and vinculin from placenta isolated from male and female control and Pl-MCUKO mice, with quantification (n = 5–8). (C–F) Seahorse mitochondrial respiration analysis using mitochondrial stress test kit from isolated primary placental trophoblast from male and female control and Pl-MCUKO mice at embryonic day 14.5, with quantifications (n = 10). Statistical analyses were conducted using 1-way ANOVA with significance ∗p < 0.05, ∗∗p < 0.01.

3.4. Transcriptomic & metabolomic analysis of murine placenta with MCU deletion

To gain deeper mechanistic insights into this model, we performed unbiased RNA sequencing on placentas (e17.5) from both sexes, with and without MCU expression. Differentially expressed genes (DEGs) and gene set enrichment analysis (GSEA) were used to evaluate the resulting transcriptomic changes. While comparison of DEGs from male and female placentas revealed sex-specific expression of X- and Y-linked genes, GSEA further identified differential regulation of interferon alpha/gamma responses, the unfolded protein response (UPR), mTORC1 signaling, and hypoxia pathways (S.Figure 3A–B). In female Pl-MCUKO placentas compared with female controls, we observed expected alterations in MCU, confirming knockout efficiency, along with changes in genes such as S6, synaptophysin, and alpha-synuclein (Figure 3A; S.C). GSEA of these samples revealed repression of G2/M checkpoint and mitotic processes, along with activation of inflammatory pathways, increased UPR, and apoptosis, suggesting that MCU deletion may impair placental growth (Fig 3B). In male Pl-MCUKO placentas relative to male controls, alterations were detected in pathways related to DNA repair, oxidative phosphorylation, and inflammatory responses, including allograft rejection (S.Fig 3D). These transcriptomic findings underscore immunological differences between male and female placentas and highlight the impact of MCU deletion on placental growth capacity.

Figure 3.

Transcriptomic and Metabolomic analysis of male and female placenta from control and Pl-MCUKO mice. (A-B) Heat map of differentially expressed gene (DEG) and gene set enrichment analysis (GSEA) from transcriptomic analysis of female Pl-MCUKO vs female control placenta collected at embryonic day 17.5 (e17.5). (C) KEGG analysis of significantly altered metabolites between female control and Pl-MCUKO placenta collected at embryonic day 17.5 (e17.5). (D) Carnitine bound lipid species that were significantly altered between female control and Pl-MCUKO placenta. (E-F) Venn diagram comparing list of significantly altered metabolites between sex and genotype.

Alongside transcriptomic analysis, untargeted global metabolomics was conducted to investigate impact of MCU deletion on metabolic alterations in the placenta. A total of 2,106 endogenous metabolites was identified and analyzed. First, we identified 216 differentially altered metabolites between control male and female placenta. KEGG analysis of these samples revealed alterations in key metabolic pathways related to linoleic acid, choline, glycerophospholipids, and arachidonic acid metabolism (S.Fig 3E). A closer examination of the differential metabolites showed a significant increase in prostaglandins and their derivatives in female placentas compared to males (S.Fig 3F). Additionally, subspecies of lysophosphatidylcholine and lysophosphatidylethanolamine were reduced in female placentas relative to males (S.Fig 3G). Interestingly, some cholesterol-derived metabolites, such as pregnenolone and 5beta-pregnane-3,20-dione, were reduced in female placentas, while 17alpha-estradiol levels were increased (S.Fig 3H). Furthermore, purine-associated metabolites and urea cycle intermediates were also altered (S.Fig 3I). Next, we identified 113 differentially altered metabolites between female Pl-MCUKO and female control placenta. KEGG analysis revealed differential metabolic pathways in biotin metabolism, protein digestion/absorption, and caffeine metabolism (Fig 3C). An inspection of the differential metabolites showed significant increases in multiple species of carnitine-bound lipids (Fig 3D), which may suggest differences in lipid metabolism following MCU deletion. Moreover, glucose levels were elevated in the Pl-MCUKO female placenta, while UDP-glucose levels were reduced, which may indicate changes in glycogen metabolism (S.Fig 3J). We also observed a trend toward increased progesterone levels in Pl-MCUKO female placentas, accompanied by decreased 17alpha-estradiol and androsterone sulfate levels, which may suggest preferential shift toward progesterone production rather than further metabolism into sex hormones (S.Fig 3K). Additionally, metabolites involved in the ammonia/urea cycle were differentially altered (S.Fig 3L). When comparing male Pl-MCUKO placentas to control male placentas, we only observed 73 differentially regulated metabolites. KEGG analysis revealed altered metabolic pathways involved in the biosynthesis of cofactors, cholesterol metabolism, and retinol metabolism (S.Fig 3M). Interestingly, when comparing the lists of differential metabolites between female and male placentas (both control and MCU knockouts), we identified 37 overlapping metabolites, including metabolites such as 5beta-Pregnane-3,20-dione, 6-keto-prostaglandin F1alpha, Xanthosine-5′-monophosphate, suggesting that these may be sex-driven, independent of MCU deletion (Fig 3E). Furthermore, comparing the differential metabolites between Pl-MCUKO and control placentas for both sexes revealed only three overlapping metabolites (Oxiglutatione, p-Cresol, Carbamoyl phosphate), indicating that MCU deletion leads to unique metabolic changes in males and females (Fig 3F). Altogether, these data suggest that MCU deletion in the placenta induces distinct metabolic profiles that may differentially impact placental function and fetal development.

3.5. MCU deletion leads to reduced fetal growth in a sex dependent manner

We next investigated whether placental MCU deletion leads to changes in placental morphology, weight, and fetal weight at embryonic day 17.5 (e17.5). The mouse placenta consists of two main fetal-derived regions: the junctional and labyrinth zones [36]. We measured the area of each zone and found no significant differences in the relative zonal areas between control and Pl-MCUKO placentas for either sex (Figure 4A). Placenta weights were similar between control and male Pl-MCUKO mice; however female Pl-MCUKO placentas were significantly smaller than their controls (Figure 4B–C). A similar trend was observed in fetal body weight, with female Pl-MCUKO offspring showing reduced weight compared to controls, while male offspring were unaffected (Figure 4D–E). No significant differences in fetal pancreas or liver weights were observed between control and Pl-MCUKO mice of either sex at this age (S.Figure 4A–B). Notably, the reduced body weight in female Pl-MCUKO fetus appeared to persist until birth (S.Fig 4C). Together, these results suggest that placental mitochondrial dysfunction is sufficient to cause reduce placental and fetal weights.

Figure 4.

Placenta MCU deletion leads to reduced fetal growth in female mice. (A) Representative H&E-stained images of placenta and relative placental zone areas from IHC analysis of male and female, control and Pl-MCUKO mice at embryonic day 17.5 (e17.5) (n = 4). White dash line represent border between junctional (Jz) and labyrinth (Lb) zones. Scale bar = 500 um. (B-E) Placenta weight and fetal body weight of male and female, control and Pl-MCUKO mice at embryonic day 17.5 (n = 7–11). Statistical analyses were conducted using 2-way student t-test with significance ∗p < 0.05.

Adult female offspring with placental MCU deficiency exhibit reduced weight gain and glucose homeostasis in HFD.

By postnatal day 7, Pl-MCUKO mice caught up in body weight, with no significant differences observed between the genotypes (Figure 5A–B). Interestingly, analysis of litter sizes at P7 showed a statistically significant reduction in female Pl-MCUKO mice and a trend toward fewer male Pl-MCUKO mice compared to controls. This may suggest potential survival issues between prenatal development and early life; however, further studies are required to confirm a survival effect (S.Figure 5A–B). At 6 weeks of age in NCD, there were no differences in body weight (Figure 5C–D), fasting glucose levels, glucose tolerance, insulin tolerance, non-fasted serum insulin level or pancreatic beta cell mass (Figure 5E–F; S.Figure 5C–H). These findings indicate that although placental MCU deletion led to reduced fetal growth, the offspring were able to undergo catch-up growth and maintain normal glucose homeostasis into young adulthood.

Figure 5.

Normal body weight and glucose tolerance in Pl-MCUKO mice. (A-D) Body weight at 7-days and 6-weeks old male and female, control and Pl-MCUKO mice (n = 7–15). (E-F) Glucose tolerance test (2 g/kg glucose) of male and female, control and Pl-MCUKO mice at 6-weeks of age. (n = 7–15). Statistical analyses were conducted using 2-way student t-test and 2-way ANOVA with significance ∗p < 0.05.

To test their ability to adapt to metabolic challenges, male and female control and Pl-MCUKO mice were fed a high-fat diet (HFD) for 20 weeks. In female mice, Pl-MCUKO mice gained less weight than controls (Figure 6A, S. Figure 6A), with no significant changes in non-fasted blood glucose levels during this period (S.Fig 6B). In terms of glucose homeostasis, Pl-MCUKO females showed a trend toward improved glucose tolerance (GTT) at 7 weeks post-HFD, with a significant improvement observed at 10 weeks post-HFD (Figure 6B, S. Fig. 6C). However, we note that fasting blood glucose levels during GTT testing were lower in Pl-MCUKO females compared to controls (Figure 6B, S. Fig. 6C). When GTT curves were normalized to fasting glucose levels, no significant differences in GTT were observed between genotypes (Figure 6C, S.Fig. 6C), which suggests that the apparent improvement in GTT may be primarily driven by lower fasting glycemia. Pl-MCUKO mice also exhibited better insulin sensitivity compared to controls (Fig 6D) at 14-weeks post-HFD. Insulin secretion following an acute glucose injection did not differ between the genotypes (S.Figure 6D–E), and there were no differences in an acute glucose-stimulated insulin secretion at the islet level (S.Figure 6F–G). However, non-fasted serum insulin levels were lower in female Pl-MCUKO than the controls (Fig 6E). And corroborating this finding, ex vivo analysis revealed a lower beta-cell mass in Pl-MCUKO females than controls in HFD (Figure 6F; S.H). Both control and Pl-MCUKO females increased their beta-cell mass under HFD; however, the control females displayed more robust fold (3.6x) increase from NCD to HFD treatment (Fig 6G), suggesting reduced capacity for beta-cell adaptation in Pl-MCUKO female mice. Because metabolic programming can be sex specific, we also tested the Pl-MCUKO males under metabolic challenge. Under HFD feeding, we observed no changes in body weight gain, glucose tolerance, insulin sensitivity, and serum insulin levels between male Pl-MCUKO and controls (S. Fig 6H-P). Together, these data highlight that FGR induced by placental mitochondrial dysfunction leads to reduced diet-induced obesity and improved glucose tolerance and sensitivity in response to metabolic challenge by HFD only in females.

Figure 6.

Female Pl-MCUKO resists diet-induced obesity. (A) Body weight of female control and Pl-MCUKO mice fed high-fat diet (HFD) over 20-weeks (n = 14). (B-D) Glucose (2 g/kg glucose) and insulin (0.5 U/kg insulin) tolerance tests at 10-week (absolute values & normalized to fasting blood glucose) and 14-week post-HFD feeding between female control and Pl-MCUKO mice (n = 8–15). (E) Non-fasted serum insulin and (F) pancreatic β-cell mass between female control and Pl-MCUKO mice at 20-week post-HFD feeding (n = 4–5). (G) Comparison of β-cell mass presented in NCD (S.Figure 5F) and HFD (Figure 6F). Statistical analyses were conducted using 2-way student t-test and 2-way ANOVA with significance ∗p < 0.05.

3.6. Reduced adiposity in female Pl-MCUKO mice under HFD feeding

To investigate the possible mechanisms behind reduced body weight gain in female Pl-MCUKO mice on a HFD (Figure 7A), we first assessed fat depot and lean mass using ECHO-MRI. Pl-MCUKO females had lower lean and fat mass compared to controls (Figure 7B–C), and the reduction in fat mass persisted even when normalized to body weight (Figure 7D–E). This finding was further supported by reduced weights of various fat depots, as well as organs such as the liver and pancreas (Figure 7F–J; S.Figure 7A–K). Despite these changes, there were no differences in food intake, either early or late into the HFD that we observed (Fig 7K-L). Notably, Pl-MCUKO females showed greater ambulatory activity during the night cycle (Fig 7M). Analysis of indirect calorimetry revealed no changes in the respiratory exchange ratio or energy expenditure in Pl-MCUKO females (Fig 7N-O; S.L-M). There were no differences in ECHO-MRI, food intake, or indirect calorimetry in male Pl-MCUKO mice compared to controls (S. Fig 8). These data demonstrate that the reduced diet-induced obesity in female Pl-MCUKO mice is primarily due to a lack of adiposity, independent of food intake or energy expenditure.

Figure 7.

Reduced adiposity in female Pl-MCUKO mice. (A-E) Body weight, ECHO-MRI parameters (lean and fat mass) between female control and Pl-MCUKO mice at 14-week post-HFD feeding (n = 14). (F-J) Organ weights between female control and Pl-MCUKO mice at 20-week post-HFD feeding (n = 14). (K-L) Average daily food intake between control and Pl-MCUKO female mice at 2-wks and 19-wks post HFD feeding (n = 4–8). (M) Average ambulatory activity between control and Pl-MCUKO female mice at 17-wks post-HFD (n = 14). (N–O) Indirect calorimetry measurements of respiratory exchange ratio (RER) and energy expenditure between control and Pl-MCUKO female mice at 17-wks post-HFD (n = 14). Statistical analyses were conducted using 2-way student t-test and 2-way ANOVA with significance ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.00001.

4. Discussion

In utero fetal development is a critical period during which metabolic programming occurs in the offspring, influencing tissue setpoints and susceptibility to non-communicable diseases such as obesity and diabetes later in life. The placenta, an organ present only during fetal life, plays a vital role in supplying essential nutrients, gases, and hormones for proper fetal development and growth. In this study, we aimed to investigate the role of placental mitochondria, a key organelle involved in energy production, cellular metabolism, and steroid hormone synthesis, and their connection to metabolic outcomes in offspring. Our findings show that placental MCU deletion leads to FGR exclusively in females, attributed in part by impaired mitochondrial respiration. Furthermore, when challenged with a high-fat diet as adults, female Pl-MCUKO mice displayed a reduction in body weight gain and improved glucose homeostasis. This study provides direct evidence that disruptions in placental mitochondrial function can drive reduced fetal growth in a sexually dimorphic manner and modulate offspring metabolic health trajectory.

In this study, we demonstrated the sexually dimorphic response to mitochondrial disruptions in the placenta. Analysis of isolated mitochondria from FGR and GDM human placentas revealed that while male placentas showed no differences under these conditions, female placentas from FGR pregnancies exhibited significantly lower activity in ETC complex II and female placentas from GDM exhibited higher activity in complexes I and IV. An important new finding is that mitochondrial respiration was enhanced in female primary mouse trophoblasts at baseline. MCU deletion resulted in reduced mitochondrial function exclusively in females, leading to poorer placental outcomes in a sex-specific manner. Sexual dimorphism is increasingly recognized as an important factor in placental development and function, with specific differences highlighting cellular metabolism and mitochondrial respiration [30,[37], [38], [39], [40]] One potential explanation for this lies in X-inactivation of specific genes. The placenta has been reported to exhibit “patchy” X-inactivation, meaning certain X-linked genes may be expressed at higher levels in female placentas [41]. For example, O-GlcNAc transferase (OGT), an X-linked gene known to modulate protein post-translational modification and regulate mitochondrial biogenesis [42,43], is more highly expressed in female than male placentas [44]. Placental deletion of OGT has been reported to alter female offspring metabolic adaptation to HFD [45,46], suggesting a connection between X-linked gene expression and mitochondrial function. Placental OGT Het females, like Pl-MCUKO, display reduced body weight gain in HFD challenge [46]. Other mechanisms at play may include fetal androgen production in utero and differential placental responses to hormones [47,48]. Additional studies are needed explore how these factors contribute to the observed sexual dimorphism in placental mitochondrial function and its impact on fetal development and health trajectory.

Mitochondrial dysfunction likely disrupts various biological processes, including bioenergetics, metabolism, and hormone synthesis. We focused on MCU, because it is a key regulator of mitochondrial calcium uptake, a co-factor essential for mitochondrial respiration, and MCU expression has been linked to placental growth and trophoblast development [23,49]. MCU and its complex expression is known to be regulated by both extrinsic and intrinsic stimuli, including nutrients (e.g. glucose, palmitate), hypoxia, inflammation, intracellular Ca²⁺ levels, and transcriptional/translational mechanisms involving miRNAs and lncRNAs, and oxidative stress [[50], [51], [52], [53], [54]]. However, the specific regulatory mechanisms of MCU in the placenta remain unknown. In this study, we show that MCU deletion in the placenta leads to reduced mitochondrial oxygen consumption rate, a proxy for ATP production. Given the placenta's rapid growth and high energy demands for growth and remodeling [6,55], these findings align with transcriptomic analysis showing altered mitosis and the reduced placental weight observed in Pl-MCUKO mice. Additionally, TCA cycle metabolites like citrate and acetyl-CoA are crucial for trophoblast development and differentiation through histone acetylation, potentially impacting epigenetics in the Pl-MCUKO placenta [56,57]. Beyond growth, placental function is closely linked to mitochondrial activity, particularly in nutrient transport, which is vital for fetal growth and metabolic tissue development. Some macronutrients and micronutrients, such as calcium and amino acids, require ATP hydrolysis for active transport or depend on ATP-driven salt gradients for co-transport [[58], [59], [60], [61]]. Mitochondria also play a key role in synthesizing steroid hormones and their precursors, such as progesterone, which are essential for pregnancy [62,63]. This process is calcium-dependent, suggesting a direct link between mitochondrial function and hormone synthesis [64,65]. Metabolomics of female Pl-MCUKO vs control placentas showed potential differences in progesterone and steroid-hormone metabolites. Future work will be focused on analyzing steroid hormone synthesis in MCU-deficient placentas to further understand the relationship between mitochondrial calcium levels and hormone production.

Pregnancy complications like FGR are often associated with increased risk of metabolic disorders in adulthood, including obesity and diabetes, due to dysregulation of metabolic tissues like pancreatic beta cell function and insulin signaling in the liver [28,[66], [67], [68], [69]]. Deletion of MCU in the placenta resulted in FGR-like phenotype in utero only in females despite successful deletion of the protein in the placentas of both sexes. However, by postnatal day 7, these mice underwent catch-up growth and showed no significant changes in body weight or glucose homeostasis, including insulin levels, when fed a normal chow diet as young adults. When exposed to HFD feeding, adult female Pl-MCUKO mice exhibited a significant reduction in weight gain and improved glucose tolerance and insulin sensitivity compared to their littermate controls. These data suggest that the reduced weight gain, specifically in response to HFD feeding, may be linked to fetal programming of lipid-metabolizing tissues, such as the liver and adipose tissue [70,71] or site of lipid absorption such as the small intestine [72,73]. Additionally, because of their sensitivity to nutrients early in life, the pancreatic beta-cells are susceptible to perturbations by maternal diet, stress and placental insufficiency [74]. While adult female Pl-MCUKO beta-cell mass was not different from control mice in normal chow diet, HFD increased beta cell mass in both the female Pl-MCUKO and control, although the magnitude of change was significantly reduced in female Pl-MCUKO. It is now well documented that insulin is required for obesity to occur [[75], [76], [77], [78]], and lack of sufficient insulin response to insulin demand may be a contributor to altered body weight gain observed in the placental MCU deficient model. Further work is required to elucidate whether reduced beta-cell mass in HFD is due to less body weight gain or the potential programming of lower beta-cell adaptive response (i.e. insulin level) to HFD caused by Pl-MCUKO in utero exposure. Additional studies, such as omics studies across key metabolic tissues such as liver, pancreas and adipose, are also required to elucidate the mechanisms of the observed placental MCU-deficiency programming of improved metabolic health in HFD. Although this model may not fully capture the complexity of FGR, which involves a variety of mitochondrial and non-mitochondrial changes influencing placental function and fetal metabolic potential, it provides direct in vivo evidence that placental mitochondrial disruptions play a significant role in offspring metabolic health.

Limitation of the study: Although Cyp19-Cre has been shown in our hands to be robustly expressed in the placenta of both male and female mice, with minimal ectopic expression in fetal tissues, the original report describing this Cre line noted moderate Cre activity in the skin, lens, and corpus callosum of the brain [33]. To our knowledge, it remains unknown whether MCU deletion in these tissues, such as corpus callosum in the hindbrain, could influence whole-body metabolism.

In conclusion, our study highlights the important role of placental mitochondrial function in determining offspring metabolic health trajectory. In utero programming represents a critical developmental window during which susceptibility to non-communicable diseases, such as obesity and diabetes, is established. Gaining a deeper understanding on the placental origin of health will inform the development of screening methods and preventive strategies for prevalent metabolic disorders.

4.1. Resource availability

Lead contact. Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Emilyn U Alejandro (ealejand@umn.edu).

CRediT authorship contribution statement

Seokwon Jo: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Grace Chung: Writing – review & editing, Project administration, Methodology, Formal analysis, Data curation. Yu-Jin Youn: Data curation, Formal analysis. Charlotte Hunt: Writing – review & editing, Project administration, Methodology, Formal analysis, Data curation. Ava Hill: Writing – review & editing, Methodology, Formal analysis, Data curation. Megan Beetch: Writing – review & editing, Methodology, Data curation. Brian Akhaphong: Writing – review & editing, Software, Formal analysis. Elizabeth A. Morgan: Writing – review & editing, Resources. Perrie F. O'Tierney-Ginn: Writing – review & editing, Resources. Sarah A. Wernimont: Writing – review & editing, Resources. Emilyn U. Alejandro: Writing – review & editing, Supervision, Resources, Project administration, Investigation, Funding acquisition.

Author's contribution statement

Conceived and developed the study, SJ. Generated, and analyzed data, assisted with manuscript preparation, and approved final version, SJ, GC, YY, CH, AH, MB, BA. Provided human placenta samples and assisted with manuscript preparation, and approved final version, EM, PO, and SW. Interpreted the data, wrote and edited the manuscript, SJ and EUA. Acquired resources and in charge of overall direction of this work, EUA. All authors contributed to the article and approved the submitted version.

Funding

This work was supported by National Institutes of Health Grant and American Heart Association (R01DK115720, R01DK136237, R56DK136293 to EUA; F31DK131860 to SJ; R25DK140753 to YY; AHA CDA1273575 to MB; 1R01HD113553-01A1 to SAW).

Declaration of competing interest

The authors declare that they have no conflicts of interest with the contents of this article.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Data availability

Data will be made available on request.