Poorly controlled diabetes mellitus alters placental structure, efficiency, and plasticity

Animal care and treatment specifications

Holtzman Sprague-Dawley rats were housed in an environmentally controlled facility at the University of Kansas Medical Center (KUMC) with lights on from 0600 to 2000 hours and allowed free access to food and water. Female rats (8–10 weeks of age) were mated with adult male rats (>3 months of age). Mating was assessed by inspection of vaginal lavages. The presence of sperm in the vagina was designated gd 0.5. A single dose of either STZ (45 mg/kg body weight) or vehicle control solution (0.1 M citrate buffer, pH 4.2) was administered via tail vein injection on gd 6.5 so as not to negatively impact embryo implantation. In some experiments, STZ and vehicle control-treated pregnant rats were placed in a hypoxic (10.5% (vol/vol) oxygen) gas-regulated chamber (BioSpherix, Lacona, New York, USA) from gd 6.5 to 9.5 or gd 6.5 to 13.5, which represent critical windows in placental development. Pregnant rats housed in ambient conditions (~21% (vol/vol) oxygen) served as controls.

Blood glucose measurements

Body weights and blood glucose were determined prior to and after treatment to ensure equal dosage and hyperglycemia, respectively. Blood glucose levels were measured from the tail vein using the OneTouch Ultra Smart blood glucose monitoring system (LifeScan, Milpitas, California, USA).13

Tissue collection

Rat pancreas, spleen, liver, and placental tissues were collected and weighed on gd 13.5 and gd 18.5. Entire placentation sites, including the ectoplacental cone, were dissected on gd 9.5. Placentation sites were dissected into junctional zone, labyrinth zone, and metrial gland on gd 13.5 and 18.5, as previously described.14 Selection of gds for analysis represented three critical stages of placentation: (1) the onset of placentation, gd 9.5, (2) establishment of placental zonation, gd 13.5, and (3) robust infiltration of invasive trophoblast cells into the uterine parenchyma, gd 18.5. Placentation sites were frozen in dry-ice cooled heptane for histological and morphometric analyses or dissected placentation site compartments snap-frozen in liquid nitrogen and stored at −80°C until processing for biochemical analyses.

Rat trophoblast stem (TS) cell culture

Blastocyst-derived rat TS cells15 were cultured in basal culture medium (RPMI 1640 (Cellgro), 20% fetal bovine serum (Sigma-Aldrich), 100 µM 2-mercaptoethanol (M7522; Sigma-Aldrich), 1 mM sodium pyruvate (ThermoFisher, 11 360–070), 50 µM penicillin (15140122; ThermoFisher), and 50 U/mL streptomycin (15140122; ThermoFisher)) supplemented with 70% rat embryonic fibroblast-conditioned medium, fibroblast growth factor 4 (25 ng/mL; Sigma-Aldrich), and heparin (1 µg/mL; Sigma-Aldrich). To model hyperglycemia, 25 mM glucose (Sigma-Aldrich) was added to the culture medium. D-mannitol (Sigma-Aldrich) was added to the culture medium to control for osmolality. Rat TS cells were exposed to ambient or low oxygen (0.5% or 1.5% O₂) tensions and 5% CO₂ at 37°C in an NAPCO 8000 incubator (ThermoFisher) for 24 hours and then harvested for immunocytochemical or biochemical analyses.

Immunohistochemistry and immunocytochemistry

Frozen placental tissues were sectioned at a thickness of 10 µm with a cryostat. Immunohistochemical analysis was performed on placental sections by immunofluorescence detection using a pan-cytokeratin primary antibody (1:250, F3418; Sigma-Aldrich) and Alexa 488-conjugated goat-antimouse secondary antibody (1:200, A11001; ThermoFisher). Nuclei were visualized with 4’,6-diamidino-2-phenylindole (Molecular Probes, Eugene, Oregon, USA). Rat TS cells were fixed with 4% paraformaldehyde (Sigma-Aldrich) for 20 min at room temperature. Immunocytochemical analysis was performed by immunofluorescence detection using a primary antibody against prolactin family 8, subfamily A, member 5 (PRL8A5, 1:20016) followed by Alexa 568-conjugated goat-anti-rabbit secondary antibody (1:200, A11001; ThermoFisher). Images were captured on a Nikon 80i upright microscope (Nikon, Melville, New York, USA) with a Photometrics CoolSNAP-ES monochrome camera (Roper, Sarasota, Florida, USA).

Morphometric measurements

Depth and extent of invasion and size of placental compartments were performed using ImageJ (National Institutes of Health (NIH)) as previously described.17 18 Briefly, the depth of intrauterine trophoblast invasion was quantified as an invasion index, defined as the distance of distal-most cytokeratin-positive cell relative to the trophoblast giant cell layer of the placenta divided by the total distance from the giant cell layer to the outer mesometrial uterine surface. The extent of intrauterine trophoblast invasion was determined as the surface area within the uterine mesometrial compartment containing cytokeratin-positive trophoblast cells. Junctional zone thickness was estimated as the area intensely staining with antibodies to pan-cytokeratin in placental cross sections. All morphometric measurements were acquired from a histological plane at the center of each placentation site perpendicular to the fetal interface of the placenta.

RNA isolation, complementary DNA (cDNA) synthesis, and transcript measurements

Total RNA was isolated from tissues using TRIzol reagent (15596018; ThermoFisher). cDNA was synthesized from 1 µg of total RNA using the High Capacity cDNA Reverse Transcription Kit (4368813; Applied Biosystems, Foster City, California, USA). Quantitative PCR (qPCR) was performed with a reaction mixture (20 µL) containing cDNA diluted five times with water and PowerSYBR Green PCR Master Mix (4367659; Applied Biosystems) using specific primer (250 nM) sequences (online supplementary table 1). Amplification and florescence detection were carried out using an ABI 7500 Real Time PCR system (Applied Biosystems) for 40 cycles (95°C, 10 min; 92°C, 15 s; 60°C, 1 min; 95°C, 15 s; 60°C, 15 s; 95°C, 15 s). Relative transcript expression was calculated by ΔΔCt method and normalized to 18S rRNA.

RNA sequencing (RNA-seq)

Transcript profiles for gd 13.5 rat junctional zone tissue isolated from control and hyperglycemic pregnancies were generated by RNA-seq as previously described.19 cDNA libraries from total RNA samples were prepared with Illumina TruSeq RNA preparation kits (Illumina, San Diego, California, USA). Barcoded cDNA libraries were multiplexed and sequenced with a HiSeq2000 DNA sequencer (100 bp paired-end reads) using a TruSeq 200-cycle SBS kit (Illumina) at the KUMC Genome Sequencing Facility. Reads from *.fastq files were mapped to the rat reference genome (Ensembl Rnor_5.0.78) using CLC Genomics Workbench 12.0 (Qiagen, Redwood City, California, USA). mRNA abundance was expressed in reads per kilobase of exon per million reads mapped. A false discovery rate of 0.05 was used as a cut-off for significant differential expression (euglycemia vs hyperglycemia). Functional patterns of transcript expression were analyzed using Ingenuity Pathway Analysis (Qiagen).

Western blotting

Placental tissues were homogenized in radioimmunopreciptation assay lysis buffer (sc-24948A; Santa Cruz Biotechnology, Dallas, Texas, USA) supplemented with Halt Protease and phosphatase inhibitor cocktail (78443; ThermoFisher). Protein concentrations were determined by the DC protein assay (Bio-Rad, Hercules, California, USA). A total of 50 µg of protein per reaction sample were separated on 4%–20% ExpressPlus PAGE Gels (M42012, M42015; GenScript, Piscataway, New Jersey, USA), transferred to polyvinylidene fluoride blotting membrane (10600023; GE Healthcare). Following transfer, membranes were blocked in 5% non-fat milk in Tris-buffered saline with 0.1% Tween 20, for non-specific binding and subsequently probed with specific primary antibodies to prolactin family 3, subfamily d, member 4 (PRL3D4, 1:50020), PRL8A5 (1:50016), actin, beta (ACTB, 1:4000, A1978; Sigma-Aldrich), and glyeraldehyde-3-phosphate dehydrogenase (1:3000, ab9485; Abcam, Cambridge, Massachusetts, USA). Immunoreactive proteins were visualized by Luminata Crescendo Western HRP substrate (WBLUR0500; Millipore, Billerica, Massachusetts, USA) according to the manufacturer’s protocol.

Statistical analysis

Statistical analyses were performed using GraphPad Prism 8 software. Welch’s t-tests, Brown-Forsythe and Welch analysis of variance (ANOVA) tests, and two-way ANOVA tests were applied when appropriate. Data are represented as mean±SD with the statistical significance level set at p<0.05.

Data and resource availability

The datasets generated and analyzed during the current study are available in the Gene Expression Omnibus website (https://www.ncbi.nlm.nih.gov/geo/; accession no GSE144276). All data generated and analyzed during this study are included in the published article and the online supplementary files. Resources generated and analyzed during the current study are available from the corresponding authors on reasonable request.

Diabetes-induced placental malformation decreases placental efficiency and fetal size

To determine the effect of diabetes on hemochorial placentation, pregnant rats were treated with STZ beginning on gd 6.5 to induce hyperglycemia post-implantation but prior to placental formation (figure 1A). This STZ model caused a decrease in maternal body weight on gd 13.5 (p<0.05), which is indicative of uncontrolled diabetes (online supplementary figure 1). There were no significant effects of STZ-induced hyperglycemia on maternal pancreas, liver, or spleen weights at gd 13.5 or gd 18.5 (online supplementary figure 1). STZ-treated rats had significantly elevated blood glucose levels compared with vehicle control-treated rats on both gd 13.5 and gd 18.5, indicating that STZ was effectively disrupting beta cell function (p<0.0001; figure 1B). Conceptus viability was not significantly different between vehicle control and STZ-treated rats on gd 18.5, despite significantly increased blood glucose levels (figure 1C). To assess the impact of hyperglycemia on placentation and fetal growth, placental and fetal weights were examined on gd 18.5, a developmental state characterized by a fully mature placentation site. Hyperglycemia was associated with significantly increased placental weight (p<0.001) and significantly decreased fetal weight (p<0.0001; figure 1D). Fetal to placental weight ratio, a measure of placental efficiency, was significantly decreased in hyperglycemic rats compared with the euglycemic control (p<0.0001; figure 1D). No significant sex differences were observed in placental weight, fetal weight, or the fetal to placental weight ratio (online supplementary figure 2) on gd 18.5. Taken together, these results suggest that exposure to STZ-induced hyperglycemia, a hallmark of poorly controlled diabetes, led to placental deficits and, as a result, diminished fetal size.

Request permission

Expanded junctional zone contributes to increased placental size

To further identify the specific placental malformations resulting from diabetes exposure, placentation sites were harvested and their histological organization were assessed using immunofluorescence staining with a pan-cytokeratin antibody. The pan-cytokeratin antibody preferentially detects trophoblast cells within the junctional zone of the placenta (figure 1E). In contrast, trophoblast-associated cytokeratin expression in the labyrinth zone is diminished relative to the junctional zone because of its substantial mesenchymal cell content. Using this immunohistochemical approach, junctional and labyrinth zone areas were quantified. Area quantification confirmed that both junctional and labyrinth zones are significantly expanded in placentas from hyperglycemic pregnancies compared with euglycemic pregnancies (p<0.001, p<0.0001; figure 1F). Expanded regions of low cell density, corresponding to a greater number of glycogen cells, fill the junctional zones of placentation sites from hyperglycemic rats and contribute to the observed increase in junctional zone area.

Diabetes alters junctional zone gene expression profiles

To gain insights into molecular mechanisms associated with the expanded junctional zone of placentas from dams with poorly controlled diabetes, we performed RNA-seq to assess transcriptomic profiles from euglycemic and hyperglycemic pregnant rats on gd 13.5 (n=4/group). By gd 13.5, placental zonation is established, but infiltration of invasive trophoblast cells into the uterine parenchyma has yet to occur. Analysis of the RNA-seq data identified 89 significantly upregulated transcripts and 319 significantly downregulated transcripts in the hyperglycemic group compared with the euglycemic group (figure 2A). Pathway analysis encompassing all significantly altered transcripts identified several cell processes implicated in junctional zone changes in response to uncontrolled diabetes. Inflammation, cell death, and glucose metabolism disorder were highly associated with the upregulated transcripts (figure 2A). Conversely, significantly downregulated transcripts in the hyperglycemia group included inflammatory response, cellular growth and proliferation, macronutrient metabolism, cell signaling, as well as cell survival and differentiation. Several significantly upregulated and downregulated genes were validated by reverse transcription quantitative polymerase chain reaction (RT-qPCR) (figure 2B). Among the differentially expressed transcripts were two prolactin family members, Prl3d4 and Prl8a5, which are known markers of junctional zone development.16 20 PRL3D4 protein levels were quantified in junctional zone tissue collected from euglycemic and hyperglycemic pregnancies on gd 13.5 and gd 18.5 by western blotting. PRL3D4 was significantly decreased in junctional zone tissues from hyperglycemic pregnancies at both gd 13.5 and 18.5 (gd 13.5, p<0.0001, gd 18.5, p<0.01; online supplementary figure 3). Diminished PRL3D4 expression was also observed within junctional zone compartments of placentation site cross sections from hyperglycemic pregnancies (online supplementary figure 3). Overall, the findings indicate that poorly controlled diabetes induces dysregulation of junctional zone development.

Request permission

Elevated glucose disrupts rat TS cell differentiation

One of the primary manifestations of diabetes, and a phenomenon we observe in our rat model, is hyperglycemia. To determine if elevated glucose impacts TS cell behavior, we cultured rat TS cells in 5 mM glucose (approximating a euglycemic state) or 25 mM glucose (approximating a hyperglycemic state) and assessed the ability of TS cells to undergo cell differentiation, focusing on junctional zone transcripts affected by hyperglycemia in vivo (figure 2B). Transcript levels were used to determine the extent of differentiation throughout a 16-day differentiation time course. In 5 mM glucose, rat TS cells begin to induce expression of genes indicative of trophoblast cell differentiation (Alox15, Prl3d4, Tpbpa, Taf7l) with increasing transcript levels observed until day 16 (figure 3A). Under TS cell-differentiating conditions in the presence of elevated glucose (25 mM), Alox15 is induced to a significantly greater extent on days 12 and 16 of differentiation compared with culture in 5 mM glucose exposure (figure 3A). In contrast, rat TS cells differentiated in 25 mM glucose exhibit significantly decreased expression of other differentiation-associated transcripts at day 16 of culture, including Prl3d4, Plac1, Tpbpa, Taf71, and Prl8a5 (figure 3A). A similar inhibition of PRL8A5 and PRL3D4 protein expression was also observed in 25 mM glucose-exposed TS cells (figure 3B,C; online supplementary figure 4). Overall, the results indicate that elevated glucose impairs TS cell differentiation and parallel in vivo placental responses to hyperglycemia.

Request permission

Diabetes impairs intrauterine trophoblast invasion

The junctional zone is prominently altered in our rat model of poorly controlled diabetes (figures 1 and 2) and represents the site of progenitors for invasive trophoblast lineage development within the rat placenta.6 21 Therefore, we sought to determine whether hyperglycemia impacts development of the invasive trophoblast lineage in vivo. To assess this, we quantified the depth and extent of cytokeratin-positive trophoblast cell invasion into the uterus and measured transcript levels of invasive trophoblast-specific markers (Prl5a1, Prl7b1, Plac1) in dissected metrial glands, the site of intrauterine trophoblast cell invasion.

Quantitation of the depth of trophoblast invasion by cytokeratin staining in euglycemic and hyperglycemic pregnant rats was assessed on gd 18.5. Endovascular trophoblast invasion depth was similar in euglycemic and hyperglycemic pregnant rats (figure 4B). However, interstitial trophoblast invasion was almost completely absent on gd 18.5 in hyperglycemic rats compared with euglycemic rats (figure 4A). Total area of trophoblast invasion, including endovascular and interstitial, was decreased in hyperglycemic compared with euglycemic pregnant rats (figure 4C). Prl5a1 and Prl7b1 are specific markers of intrauterine invasive trophoblast cells,22 23 and Plac1 has recently been implicated as an intrinsic regulator of the invasive trophoblast lineage.24 Prl5a1, Prl7b1, and Plac1 transcript levels were significantly decreased within the uterine–placental interface of hyperglycemic versus euglycemic pregnancies (figure 4D). Collectively, the results indicate that hyperglycemia adversely affects development of the invasive trophoblast lineage.

Request permission

Hyperglycemia and placenta plasticity

A fundamental property of the placenta is its plasticity.25 The placenta possesses the ability to adapt to a wide range of environmental challenges that could threaten pregnancy.6 25 Adjustments in trophoblast–uterine spiral artery interactions are a key component of placental adaptations to environmental stressors.4 26 Hypoxia has proven to be an effective tool to elicit adaptations in the development of the maternal–fetal interface.17 Since hyperglycemia affected both junctional zone and invasive trophoblast cell lineage development, we sought to determine the impact of hyperglycemia on hypoxia-induced placental plasticity.

We used an experimental paradigm of hypoxia exposure in the pregnant rat proven to activate endovascular trophoblast invasion and trophoblast-guided uterine spiral artery remodeling.17 Euglycemic and hyperglycemic rats were exposed to ambient conditions (~21% oxygen) or hypoxia (10.5% oxygen) beginning on gd 6.5 and sacrificed on gd 13.5 (figure 5A). The extent of STZ-induced hyperglycemia was similar in ambient and hypoxia exposed pregnant rats (figure 5B). Hypoxia exposure significantly decreased uterine + conceptus weight in hyperglycemic compared with euglycemic pregnant rats at gd 9.5 (p<0.01; figure 5A) but did not adversely affect conceptus viability. Conceptuses from euglycemic pregnant rats exposed to ambient or hypoxic conditions and conceptuses from hyperglycemic pregnant rats exposed to ambient conditions survived to gd 9.5 (figure 5D). However, by gd 13.5 all hyperglycemic rats exposed to hypoxia showed complete pregnancy loss (p<0.0001; figure 5E). These findings prompted examination of conceptus development in each group at an earlier stage of gestation (gd 9.5). Histological analysis of gd 9.5 conceptuses demonstrated that development of the primordial placenta, referred to as the ectoplacental cone, was adversely affected in hyperglycemic pregnancies exposed to hypoxia (figure 5F), which was verified by the quantification of ectoplacental cone surface area (p<0.0001; figure 5G) and diminished proliferation, as measured by phosphohistone H3 immunoreactivity (p<0.0001; figure 5H). Thus, hyperglycemia adversely affects placental adaptations to hypoxia.

Request permission

High glucose blunts TS cell responses to low oxygen

To determine whether the interaction of hyperglycemia and hypoxia was directly altering trophoblast lineage development, we investigated the impact of glucose on TS cell responses to low oxygen in vitro. Transcript levels of known TS cell hypoxia-responsive genes, including IL33 and CD200, were measured by RT-qPCR.27 IL33 and CD200 transcript concentrations were similar in TS cells cultured in 5 mM or 25 mM glucose exposed to ambient conditions (p<0.01; online supplementary figure 5). In contrast, exposure to low oxygen significantly induced expression of IL33 and CD200 in 5 mM glucose (p<0.01) but IL33 and CD200 were not induced by hypoxia in 25 mM glucose (p<0.01). We also examined the interactive effects of glucose and hypoxia on TS cell number. TS cell viability was not altered by the culture conditions (figure 5I); however, low oxygen stimulated TS cell proliferation in 5 mM glucose, but not in 25 mM glucose (p<0.0001; figure 5I). These in vitro results support the in vivo observations that diabetes (hyperglycemia) alters placental cell plasticity through, at least in part, direct actions on trophoblast cell development.