This protocol received approval from the Biosafety, Ethical, and Research Committee of the Instituto Nacional de Perinatología (INPer, code numbers 2023-1–6 and 2018-1-152), as well as from the Hospital Ángeles México (HAMx-MPVS-2023-1), Hospital Ángeles Lomas (HAL 366/2020), all located in Mexico City. Before the cesarean section, written informed consent, as outlined in the Declaration of Helsinki, was obtained voluntarily from each mother.

DMEM low glucose culture media (#BIO-L0066-500) and newborn calf serum (#S0750-500) were from Biowest (Riverside, MO, USA). Monohydrated dextrose (#JT1910-1) and D-mannitol (#2554-01) were acquired from JT Baker- Fisher Scientific (Mexico City, Mexico) and were diluted in deionized water to 1M solution, filtered, and preserved at 4 °C. Dextrose (D-glucose) was supplemented to DMEM media to obtain a glucose curve of 5 mM (control glucose), 10 mM and 50 mM (high glucose concentrations). Mannitol was also supplemented in culture media at 45 mM (osmotic control). Lipoteichoic acid (LTA, #L215) from Staphylococcus aureus was from Sigma-Aldrich (St. Louis, MO, USA). Streptococcus agalactiae Lehmann and Neumann was obtained from the American Type Culture Collection (ATCC) #27956. (Rockville, MD, USA). Blood agar (#7242) was obtained from MCD Lab (Mexico). Sterile defibrinated ram blood (#1600 FC) was from Dibico (Mexico).

We collected complete placenta specimens from healthy term (37–41 weeks) pregnant women carrying a single fetus, scheduled for cesarean section at the mentioned hospitals in section 2.1.

We excluded from the study patients who had experienced urinary or cervicovaginal infections during the last trimester of pregnancy, those with allergies to penicillin or streptomycin, who had undergone vaginal delivery, with multiple pregnancies, or those with any documented metabolic or endocrine diseases. Additionally, patients with hyperglycemia during surgery (fasting blood glucose > 95 mg/dL) were excluded. Samples indicating recent infection (positive IgM) after rapid SARS-CoV2 antibody testing (placental blood drop) were also eliminated. Cesarean sections were clinically indicated based on medical records for reasons such as breech presentation, cephalopelvic disproportion, previous uterine surgery, history of myomectomy, or the personal choice of the mother.

To ensure tissue sterility, cotyledon explants were processed by standard microbiological testing methods to detect aerobic and facultative anaerobic microorganisms. Only the data from samples with known sterility were analyzed.

The procedure for isolation and culture of villous explants has been described elsewhere (25). Briefly, clean cotyledons were cut into small explants of 3 to 5 mm³, and kept in a Petri dish with saline and medium. For the experiments, explants were incubated in 24-well plates with low glucose DMEM supplemented with 10% newborn calf serum, 1% sodium pyruvate, 100 U/mL penicillin, and 100 µg/mL streptomycin, under standard conditions (37 °C, 95% relative humidity and 5% CO₂).

Villous explants were incubated for 48 hours in supplemented media adjusted to glucose 5, 10, or 50 mM based on previous experiments developed in trophoblasts and placental tissue to mimic normal, and hyperglycemic environments associated with GDM or diabetes (5, 7, 26, 27). Additionally, some explants were cultured with 5 mM glucose and 45 mM mannitol as an osmotic control, given that mannitol is a non-metabolized polyol.

At the end of the experiments, culture media were recovered and kept at -40°C for subsequent use in migration assays or analysis by the multi-analyte flux technique or ELISA. Villous explants were stored in formalin for Brown-Hopps staining, in TRIzol for RNA isolation, or RIPA buffer supplemented with protease inhibitors for the quantification of TLR-2 protein.

After 48 hours of incubation with the treatments described in section 2.5, the explants were infected with GBS (1x10⁵ CFU/mL) for 8 hours. Infection was performed in normoglycemic media (i.e. 5 mM) without antibiotics. Therefore, the bacteria were not exposed to high glucose.

At the end of the infection period, bacterial growth was quantified using the plate dilution technique, and the culture media were collected for cytokine and chemokine quantification using the multi-analyte flux technique. For the plate count, 100 μL of the supernatants were collected, and serial dilutions were performed in saline solution + 0.05% Tween 80. Each dilution was plated in triplicate (10 μL) on blood agar plates, which were incubated for 24 hours at 37°C. After incubation, β-hemolytic activity was corroborated, and the number of colonies grown in each dilution was counted. The experiment was discarded if any abnormal colony morphology was observed. The final count was expressed as CFU/mL/gram of tissue.

The tissue was embedded in paraffin blocks, and sections were cut using a microtome. The sections were then placed on slides and dewaxed. The slides were immersed in a 1% crystal violet solution for two minutes. Gram’s iodine (the mordant) was added for 5 minutes, followed by rinsing with distilled water. The slides were destained with an acetone-alcohol solution for 10 seconds. Next, 0.5% basic fuchsin was applied for five minutes, and the slides were rinsed again. A 0.1% picric acid solution was added until a slight yellow color appeared, and the slides were quickly rinsed with xylene. Finally, the stained slides were coverslipped.

Images were obtained using a Carl Zeiss Lab A.1 microscope equipped with an AxioCam Erc5s camera (Carl Zeiss, Inc., Thornwood, NY, USA) at 10X magnification. Digital processing of the images was performed with Zen 2.3 (blue edition) software (Jena, Germany). Syncytial disruption was quantified in three slides per treatment across three independent GBS-infected experiments. The physical separation of syncytia from the villous tree was recorded by three independent observers using the ImageJ 1.54g program.

After thawing the media, the secretion of cytokines (IL-1β, INF-α2, INF-γ, TNF-α, IL-6, IL-10, IL-12p70, IL-17a, IL-18, IL-23 and IL-33) and chemokines (MCP-1, RANTES, IP-10, ENA-78, Eotaxin, GRO-α, IL-8, I-TAC, MIG, MIP-1α, MIP-1β and MIP-3α) into the culture medium was quantified using commercial LegendPlex™ kits from BioLegend (#740809 and #740985, respectively. San Diego, CA, USA). The assays were performed according to the manufacturer’s instructions. The concentration of each analyte was normalized per gram of wet tissue.

After infection, the tissue was placed in TRIzol and homogenized using a polytron. It was then centrifuged at 28,000 g for 30 seconds, and the supernatant was collected. The RNA was extracted using the commercial Direct-zol RNA mini prep kit from Zymo Research (#R2050). In all cases, the amount and quality of RNA were estimated spectrophotometrically at 260/280 nm (Nanodrop. Thermo), and a constant amount (500 ng) of RNA was reverse transcribed using the commercial Maximum First Strand cDNA Synthesis Kit from Thermo-Scientific (#K1642). The qPCR amplifications were performed in identical conditions for all genes, using the commercial kit LigthCycler^® TaqMan^® Master Mix from Roche (#04535286001), loading 1 μL of cDNA. GAPDH was used as house-keeping gene to normalize gene expression. The cDNA was then submitted to standard qPCR conditions using the LightCycler 480 Instrument (Roche Diagnostics, Mannheim, Germany). The primers were synthesized by Integrated DNA Technologies. The sequence of primers and the hydrolysis probes are shown in Supplementary Table 1 .

With the assistance of surgical tools, the chorioamniotic membranes and umbilical cord were removed. Cuts were made on the maternal side of the placenta using a scalpel blade, and the placenta was transferred to a sterile bag with the maternal side facing downward. The placenta was kept in the bag for 10 minutes to allow intervillous blood to efflux by gravity. Blood was collected with a sterile transfer pipette and placed into heparin tubes. The heparinized blood was transferred to 50 mL conical tubes and diluted 1:3 with 1X Hank’s balanced solution, supplemented with 10 mM HEPES and 34.2 mM NaHCO₃.

Next, a 15 mL bed of Lymphoprep™ (STEMCELL Technologies #07851) was placed in another 50 mL conical tube. Then, 30 mL of diluted blood was gently added to the Lymphoprep bed. This tube was centrifuged for 40 minutes at 400 g at room temperature. The ring containing IVMC was then recovered. The cells were washed with 1X Hank’s solution and centrifuged at 550 g for 10 minutes. To eliminate erythrocyte contamination, the cells were incubated for 5 minutes in erythrocyte lysis buffer (NH₄Cl [155 mM], KHCO₃[10 mM], and EDTA [0.1mM]). After 5 minutes, the reaction was stopped by adding 1X Hank’s solution, and the cells were centrifuged at 550 g for 10 minutes. The resulting pellet was washed two more times with 1X Hank’s solution. The IVMC were then resuspended into RPMI-1640 medium and maintained at 37°C with 5% CO₂ until used in the functional migration assay.

After isolation of IVMC, cells were labeled with the fluorochrome CellTracker™ Green CMFDA at a 1:1000 dilution (Invitrogen #C2995). The cells were incubated in RPMI medium for 45 minutes at 37°C with 5% CO₂ in the dark. After incubation, the cells were centrifuged for 5 minutes at 550 g, and three wash cycles with Hank’s solution were performed. The cells were then resuspended in RPMI-1640 medium.

The Boyden chamber model was used with a 5 µm porous membrane. 20% newborn calf serum served as a positive chemoattraction control. 600 μL of villous-conditioned media was placed in the lower chamber as the chemotactic agent, and 500,000 IVMC in 300 μL of RPMI medium were placed in the upper chamber. The migration assay was performed for two hours at 37°C with 5% CO₂. Afterward, 150 μL from the lower chamber were collected in duplicate and placed in a 96-well plate for reading on a fluorescence spectrophotometer (Synergy 2 microplate reader, Biotek) at 492/517 nm. A calibration curve was constructed with a 7-point fold serial dilution, ranging from 125,000 to 1,953 cells. A control well of spontaneous migration, with fresh DMEM media, was also included to account for gravity-driven movement. The calculation of migrating cells was determined by subtracting the number of spontaneous migration cells from the total cells in the lower chamber.

In parallel to the previous section, after 2 hours of the IVMC migration assay, the supernatant from the lower chamber was collected and centrifuged at 400 g for 5 minutes. The cells were then labeled with titled volumes of fluorochrome-conjugated antibodies: anti-CD3-FITC, anti-CD16-PE, anti-CD56-PE, anti-CD19-APC (BD Biosciences #340500), and anti-CD14-PE-Cy7 (BD Biosciences # 663195). The labeled IVMC were incubated at room temperature for 15 minutes, protected from light.

Following incubation, the cells were fixed with 250 μL of 1X FLS solution (BD FACS™ Lysing Solution 10X concentrate. BD Biosciences #349202) and incubated for an additional 10 minutes, also protected from light. Afterward, the cells were washed with phosphate-buffered saline (PBS) and centrifuged at 400 g for 5 minutes. The supernatant was discarded, the cells were resuspended in 100 μL of PBS, and the samples were analyzed using a flow cytometer or stored at 4°C, protected from light, for up to 3 days. The FACSAria™III cytometer (BD Biosciences) was used, and 25,000 events were analyzed per sample. Unique events in the mononuclear cell region were selected and filtered based on CD14+, CD3+, CD19+, CD16+CD56+ or CD3+CD16+CD56+ expression for the detection of monocytes, T, B, NK, and NKT cells, respectively. Supplementary Figure 1 shows the analysis for the batch with selection strategy, parameters, and representative plots.

After 8 hours of infection with GBS, the villous explants were collected and placed in RIPA lysis buffer with protease inhibitors (Roche #11836170001). Samples were then stored at -40°C until mechanical lysis. The samples were lysed using a polytron (OMNI International, Tissue Master 125), and centrifuged at 2850 g for 10 minutes at 4°C. The supernatant was collected, and the protein concentration was quantified by the Bradford method. TLR-2 levels were quantified using an ELISA kit from Invitrogen (#EH459RB) according to the manufacturer’s instructions. Each sample was adjusted to 25 μg of protein in 100 μL per well. Detection range was from 0.328–80 ng/mL. For detection, 100 μL TMB (3,3′,5,5′-Tetramethylbenzidine) was used as the substrate, and 50 μL of 0.2N H₂SO₄ was added as the stop solution. Samples were read at 450 nm without correction.

All statistical analyzes were performed with GraphPadPrism 10.1.1 software (GraphPad Software Inc). A p-value of less than 0.05 (p < 0.05) was considered a statistically significant difference. Using the Shapiro-Wilk test, the distribution of the data was determined as normal or non-normal, and consequently the parametric or non-2parametric statistical test was chosen for comparison between groups. The statistical test and the post-hoc used are indicated in the figure captions.

After 8 hours of infection, we observed a significant increase in the colony-forming unit (CFU) counts of GBS in the culture media of villous explants pre-exposed to high and severely high glucose concentrations, compared to explants cultured in control medium ( Figure 1A ). Additionally, the bactericidal ability of villous cells was evident when comparing bacterial growth in the absence of explants (1.6 x 10⁹ CFU/mL/g of tissue) to bacterial growth in the presence of normoglycemic explants (1.1 x 10⁶ CFU/mL/g of tissue) ( Figure 1A ).

Next, we examined the morphological changes in villous explants after GBS infection and evaluated whether high glucose conditions affect this response ( Figures 1B–H ). Brown-Hopps staining was used to visualize bacteria, which appeared blue. GBS was primarily contained at the syncytial barrier in all glucose environments. However, higher glucose concentrations significantly disrupted the continuity of the syncytial layer ( Figure 1B ), showing signs of syncytium shedding and destruction of villous integrity in some regions (arrows and magnified insets).

Together, these findings highlight the critical impact of a high glucose environment on both the control of bacterial growth and the integrity of the syncytial barrier, which represents a physical defensive mechanism against GBS in the human placenta.

The next step was to measure the secretion profile of chemokines and cytokines. TNF-α, MCP-1, and MIP-1β ( Figure 2A ) were primarily induced by GBS infection, but the dual challenge with high glucose significantly reduced their secretion. In contrast, MIP-3α, IP-10, eotaxin, IL-6, IL-1β, and IL-10 ( Figure 2B ) although also induced by GBS infection, were not downregulated by the pre-exposure to high glucose. Instead, the secretion of these cytokines was further increased by the dual inflammatory challenge. This pattern is insightful, as it sheds light on the specific role of each molecule in modulating the immune response. For instance, as MCP-1 and MIP-1β are closely involved in recruiting monocytes and NK cells, our results highlight the repressive effect of hyperglycemia upon the chemotactic activity of these cytokines/chemokines. Another interesting result was IL-10 secretion in the absence of GBS infection. As depicted in Figure 2B , IL-10 was the only cytokine significantly upregulated by 50 mM glucose. Considering the anti-inflammatory role of this cytokine, its stimulation by high glucose might explain the absence of an increase in the secretion of inflammatory molecules under this condition as early as 8 h.

We also quantified the secretion of additional chemokines (IL-8, I-TAC, MIG, MIP-1α, TARC) and cytokines (IL-12p70, IL-17a, IL-18, IL-23, IL-33, IFN-α2), which were exclusively upregulated by GBS infection but were not further modulated by high glucose ( Supplementary Figure 3 ).

To corroborate the previous findings, we evaluated the regulation of some chemokines and cytokines at the transcription level. We observed that the dual stimulus of infection and hyperglycemia downregulated the gene expression of MCP-1 and TNF-α ( Figure 3A ), consistent with the results from the secretion assay. In contrast, IL-6 was upregulated under the dual stimulus compared to explants cultured in control glucose media and infected with GBS ( Figure 3B ). Notably, in the absence of GBS, the gene expression of MCP-1 was significantly downregulated by high glucose, while that of IL-6 was upregulated ( Figures 3A, B ). Finally, GBS infection increased IL-1β and IL-8 mRNA levels, but these were not further modulated by high glucose concentrations ( Supplementary Figure 4 ), in a similar pattern as the results from the secretion assay.

These results confirm that the combined stimulus of hyperglycemia and GBS infection led to a chemokine/cytokine-specific pattern in human villi, leading to an inflammatory chemokine imbalance.

After observing that certain cytokines and chemokines were upregulated in the combined challenge of GBS infection and hyperglycemia ( Figures 2B , 3B ), while TNF-α, MCP-1 and MIP-1β were downregulated under these conditions ( Figures 2A , 3A ), we investigated whether the chemoattraction of immune cells would be affected.

Our results demonstrated that GBS infection enhanced IVMC migration compared to uninfected groups. However, when the villous were cultured under hyperglycemic conditions, IVMC migration decreased significantly ( Figure 4A ). Next, we examined the phenotype of these migrating cells. We found that the cell populations exhibiting reduced migration were monocytes and NK cells ( Figures 4B, C ), which correspondingly are chemoattracted by MCP-1 and MIP-1β.

On the other hand, T lymphocytes and NKT cells did not exhibit migration toward villous media, regardless of high glucose concentrations or GBS infection ( Supplementary Figures 5A, C ). B lymphocytes showed a trend toward decreased migration under the combined stimulus; however, this difference was not statistically significant ( Supplementary Figure 5B ).

After observing decreased cytokine and chemokine molecules, and migration of total IVMC and specific cells, by both, hyperglycemia and infection, we sought to understand the mechanism underlying this phenomenon. TLR-2 is the main PRR responsible for sensing gram-positive bacteria such as GBS. In addition, TLR-2 signaling triggers NF-κB and regulates the expression of pro-inflammatory cytokines and chemokines (28). Therefore, we decided to study whether its expression at the protein or mRNA level is affected.

Our results showed that TLR-2 protein levels were significantly induced by high glucose levels and GBS infection. But GBS infection in explants pre-exposed to glucose 50 mM significantly diminished TLR-2 levels ( Figure 5A ). Coincidently, gene expression assay indicated that combination of GBS and high glucose levels diminished TLR-2 mRNA levels ( Figure 5B ).