This study was carried out with the approval of the Ethical Committee Nordwest und Zentralschweiz, Basel, Switzerland (EKBB 195/13 or EKNZ PB 2016-00611), with written informed consent from all subjects in accordance with the Declaration of Helsinki.

Pregnant women (n = 19: n = 10 healthy pregnant and n = 9 GDM pregnancy) were recruited at the time of their routine examination for glucose intolerance between the 24th and 28th weeks of pregnancy (median gestational age: 25 weeks and 6 days; median age: 34.5 years). Healthy non-pregnant controls matched for age (n = 10) were recruited at the Blood Bank of the Swiss Red Cross, Basel. Human placentae were obtained from pregnant women who delivered healthy, singleton infants at term (>37 weeks’ of gestation) undergoing elective cesarean section.

Inclusion criteria for the pregnant subjects included healthy singleton pregnancies. Exclusion criteria included maternal diseases like hypertension, diabetes mellitus and chronic disease, known infection like hepatitis or human immunodeficiency virus, maternal history of hypertensive diseases in previous pregnancy, and fetal genetic, chromosomal or intervention-requiring morphologic abnormalities. Inclusion criteria for non-pregnant controls were fair general condition, female sex, age ≥25 and ≤45 years and for blood donors fulfilling national criteria for blood donation. Exclusion criteria were current or previous systemic autoimmune disease, asthma, convalescence after major illness, surgery, current medication with corticosteroids, immunosuppressive agents, and malignant neoplasia or chemotherapy within 5 years before recruitment for the study.

Blood samples from patients affected by TD1 and TD2 were collected during routine visits and directly processed for analysis.

Whole blood was collected into EDTA- and silicone-coated tubes (Sarstedt) and analyzed by Hemavet 950FS (Drew Scientific) for complete blood cell counts. Plasma and serum were collected and processed as described previously (24, 33, 38). Samples were studied immediately or stored at −80°C until analyzed.

Five placentae samples for each group were collected and dissected from the middle cross section of villus tree within 15 min of the delivery. To compensate for intra-placental variability, we collected 3 independent samples per placenta giving a total of 15 GDM and 15 control samples. Tissue was bluntly dissected to remove visible connective tissue, blotted dry on filter paper, snap frozen in liquid nitrogen, and stored at −80°C. Small pieces of tissues were embedded in optimal cutting temperature compound (Tissue-Tek; Sakura Finetek USA) and kept at −80°C until cryosection.

Pregnant women between the 24th and 28th weeks of gestation, after 10–16 h of fasting, gave blood, and the fasting glucose levels were tested. They ingested 300 ml of Accu-Chek Dextrose O.G.-T. (Roche), which contains 75 g of glucose, and gave blood after 60 and 120 min, according to manufacturer’s instructions. Glucose was measured in the certified laboratory of the University Hospital Basel.

Neutrophils were isolated by Dextran-Ficoll density centrifugation (20, 40). Cell viability was assessed by trypan blue dye exclusion in a hemocytometer and was routinely 96–98% with a purity of over 95%. Neutrophils were directly seeded in 24-well or 96-well plates and allowed to settle for 15 min at 37°C under 5% CO₂ prior to further experimentation.

BeWo cells (ATCC CCL-98) were grown at 37°C under a humidified 5% CO₂/95% air atmosphere in F-12K Medium containing 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin. Experiments were performed under serum-starved conditions using cells at 80% confluence between passages 30 and 40. BeWo cells were grown under low glucose (7 mM) or HG (25 mM) conditions. Supernatants were harvested, submitted to centrifugation, aliquoted, and frozen at −80°C until use. The coculture experiments were performed in 6-well cell culture plates with 12 mm coverslips (Corning^® BioCoatTM Coverslips). When indicated, neutrophils were directly added to the BeWo cells at a density of 0.5 × 10⁶ neutrophil/ml for 3 h. In neutralization studies, 5 µg/ml of chimeric anti-TNF-α IgG1 antibody infliximab (Remicade, MSD) was given simultaneously to neutrophil addition.

Neutrophil extracellular traps were quantified by SytoxGreen fluorimetry (20, 40). Neutrophils (2.5 × 10⁴) freshly isolated were cultured in the presence of 0.2 µM SytoxGreen (Invitrogen, Life Technologies) in a 96-well dark microtiter plate at 37°C under 5% CO₂ and left untreated or stimulated with the indicated agents over 3 h. PMA was used as the positive control. Fluorescence (excitation 485 nm, emission 535 nm) was measured in a Biotek Synergy H1 Hybrid Reader (Biotek) and results given as mean DNA fluorescence intensity. Photomicrographs in bright field and green fluorescence spectra were assessed with an Olympus IX50 inverted fluorescence microscope coupled to an Olympus XM10 monochromatic CCD camera and analyzed with the Olympus CellSens Dimension software (Olympus).

Neutrophil extracellular traps were quantified by IHC staining of 2.5 × 10⁴ neutrophils per well in a 96-well plate, mouse anti-human MPO antibody (1:750, ab25989, Abcam) and rabbit anti-human citH3 antibody (1:150, ab5103, Abcam), or the respective isotype controls, followed by incubation with goat anti-mouse IgG AF555 (1:1,000, A28180, Invitrogen Life Technologies) and goat anti-rabbit IgG AF488 (1:1,000, A11034, Invitrogen Life Technologies) (24, 38). DNA was counterstained with 4′,6-diamidino-2-phenylindole (DAPI, D9542, Sigma-Aldrich). NETs were visualized by using an Olympus IX81 motorized epifluorescence microscope (Olympus) in conjunction with an Olympus XM10 monochromatic CCD camera (Olympus) and analyzed with the Olympus CellSens Dimension software (Olympus). A minimum of 20 fields at 10× magnification (at least 500 to 1,000 neutrophils) per sample were evaluated for MPO/citH3 and DNA co-staining through ImageJ analysis software (NIH); nuclear phenotypes and NETs were determined, counted, and expressed as percentage of the total area of cells in the fields. For staining and quantification of NE, 1 × 10⁵ neutrophils were seeded on poly-l-lysine-coated glass coverslips (BD Biosciences) in 24-well tissue-culture plates, fixed with 4% paraformaldehyde, and blocked overnight (HBSS with 10% FBS, 0.1% Tween20, and 2 mM EDTA) at 4°C. Samples were stained with rabbit anti-NE (1:200, ab26154, Abcam) following incubation with goat anti-rabbit IgG AF488 (1:1,000, Invitrogen Life Technologies). DNA was stained with DAPI (1:10,000, Sigma-Aldrich). Coverslips were visualized using an Olympus BX61 Diana fluorescent microscope.

For double immunohistochemistry of the human placentas, 8-µm thick placental sections were fixed in 4% paraformaldehyde, permeabilized with 0.2% Triton X-100, and blocked in 3% BSA overnight. Primary antibodies used were: rabbit anti-citH3 (1:200, Abcam), rabbit anti-MPO (1:200, Abcam), mouse anti-NE (1:200, HM2174, Hycult Biotech), and rabbit anti-IRS1 (1:200, sc-559, Santa Cruz). Sections were incubated in a light-protected humidified chamber with Alexa 488-conjugated goat anti-rabbit (1:1,000, Life Technologies) and Alexa 647-conjugated goat anti-mouse (1:1,000, A21237, Life Technologies) secondary antibodies. Sections were stained with DAPI (1:10,000) and mounted in Mowiol (Sigma-Aldrich). In addition, a further consecutive section was stained using H&E for comparison against immunostained tissues. The slides were observed under an Olympus BX61 Diana fluorescent microscope and photographed.

The cells, following coculture experiments, were fixed in 2% paraformaldehyde and blocked with 3% BSA. Primary antibodies used were: rabbit anti-citH3 (1:150, Abcam) and mouse anti-NE (1:150, Hycult Biotech), or the respective isotype controls, followed by incubation with goat anti-mouse IgG AF555 (1:1,000, A21424) and goat anti-rabbit IgG AF488 (1:1,000, A11034, Invitrogen Life Technologies). DNA was counterstained with DAPI (1:10,000).

Histone/DNA complexes in sera and plasma were measured using the Human Cell Death Detection ELISA^PLUS (Roche Diagnostics) (24, 38). The concentrations of NE in sera and plasma were measured by sandwich ELISA, utilizing the Elastase/a1-PI Complex ELISA Kit (Calbiochem) (24, 38). For the detection of MPO in sera and plasma, the Hycult Biotech ELISA Kit was used (24, 38). For the detection of A1AT in sera and plasma, the AssayPro ELISA Kit was used. TNF-α concentrations in sera and plasma were assessed by sandwich ELISA with Human TNF-α DuoSet ELISA Kit (R&D Systems).

For in vitro incubation studies, 2.5 × 10⁴ neutrophils from healthy women were treated with 6% plasma derived from non-pregnant controls, pregnant donors, and donors with GDM during second and third trimesters of gestation (24, 38). All experiments were carried out over 3 h in four to six replicates.

For similar in vitro incubation studies, 2.5 × 10⁴ neutrophils from healthy women were treated with supernatants from BeWo cells after exposure to normal or high glucose for 12 h.

To neutralize TNF-α, pooled plasma from the study groups of interest or BeWo supernatant was pretreated for 30 min with Infliximab (5 µg/ml).

For the two-step stimulation in vitro experiments, neutrophils from healthy controls were pretreated or not with TNF-α (50 ng/ml) for 30 min and then exposed to HG (25 mM) for 150 min, for a total time of 3 h.

Total cellular protein was extracted using Pierce RIPA Buffer (Thermo Scientific) supplemented with PhosSTOP Protease Inhibitor Cocktail Tablets (Roche Life Science). For the placenta, frozen tissues were thawed for several minutes in pre-chilled RIPA buffer and homogenized with a Polytron (Kinematica). Western blotting was performed by using AnykD Mini-PROTEAN TGX Gels (Biorad, Hercules, CA, USA) and nylon/nitrocellulose membranes (Biorad). Primary and secondary antibodies used were: rabbit anti-IRS1 (1:500, sc-559, Santa Cruz Biotechnology), rabbit anti-GLUT4 (1:1,000, 07-1404, Millipore), mouse anti-β-Actin (1:5,000, Sigma-Aldrich), goat anti-Mouse (sc-2005, Santa Cruz Biotechnology), and/or anti-Rabbit (sc-2004, Santa Cruz Biotechnology), human anti-HRP (1:2,000, Santa Cruz Biotechnology). HRP activity was detected by using SuperSignal West Pico Chemiluminescent Substrate (Thermo Scientific). Densitometric analysis and protein quantification of the western blots was performed by using Image Lab Software (version 5.2.1). Blots were stripped briefly in 0.2 M sodium hydroxide and reprobed for mouse anti-β-Actin (Sigma-Aldrich) as a loading control.

Total RNA was isolated from 3 × 10⁶ neutrophils by using the RNeasy Mini Kit (Qiagen). TaqMan real-time quantitative RT-PCR was performed utilizing the Applied Biosystems StepOne Plus cycler (Applied Biosystems) and TaqMan Gene Expression Assay primer and probe sets (Applied Biosystems) for ELANE (HS00236952_m1), TNF (HS01113624_g1), and RPLP0 (HS99999902_m1). Data were normalized to the housekeeping gene B2M (HS99999907_m1), after a selection procedure from six different endogenous reference genes, as suggested in the MIQE guidelines. Relative values were calculated with 2^−DDCt analysis.

BeWo cells were preincubated with 80 nM of human NE (SE563, Elastin products company) in glucose-free KrebsRinger HEPES buffer (KRH) [25 mM Hepes-NaOH (pH 7.4), 120 mM NaCl, 5 mM KCl, 1.3 mM CaCl₂, 1.3 mM KH₂PO₄] for 3 h. Then, cells were rinsed with KRH and stimulated in the same buffer containing 100 nM of insulin (Actrapid HM) for 1 h. Following stimulation, the cells were washed and in order to measure the glucose uptake, incubated in KRH buffer supplemented with 1 μCi/ml of 2-[1,2-3H(N)]-deoxy-d-glucose (Perkin Elmer) for 10 min. Resulting cell lysate was mixed with scintillation fluid (Filter-count, Perkin Elmer), and radioactivity was measured in a liquid scintillation analyzer (Packard 1900 TR). Glucose uptake was expressed in triplicate, and the mean SD of the data was calculated.

All data are presented as mean ± SEM. Descriptive statistics for continuous parameters consisted of median and range, and categorical variables were expressed as percentages. Comparisons between patients and healthy controls were carried out by the Mann–Whitney U test with a Welch post-test correction. Statistical significance in multiple comparisons was by one-way analysis of variance (ANOVA) with a Dunn’s post-test correction. P values <0.05 were considered statistically significant. Data were processed in GraphPad Prism version 6.0 for MacOSX (GraphPad Software Inc.).¹ Andreas Schoetzau provided professional statistical assistance.²

In an exploratory examination of the NETotic activity of freshly isolated neutrophils in vitro, we observed that NET formation was significantly elevated in a case with GDM when compared to a matching second trimester sample (Figure 1A). This feature was confirmed in a larger series of GDM cases (n = 9) and matching healthy control pregnancies (n = 10) and non-pregnant donors (n = 10) (Table S1 in Supplementary Material; Figure 1B). In these analyzes, it was noted that GDM samples contained more neutrophils activated to undergo NETosis as apparent from the presence of citrullinated histone H3 (citH3 +) moieties (Figure 1B), a key feature in the NETotic signaling cascade. This was mirrored in a greater tendency to undergo NETosis in vitro, as indicated by the quantity of extracellular NE/DNA (total area) (Figure 1B, lower panel). Furthermore, this NETosis increased from the time of diagnosis (24–28 weeks of gestation) throughout the second and third trimesters of gestation as measured with cell-free nucleosomes in the supernatant of freshly isolated neutrophils (Figure S1 in Supplementary Material).

Since numerous reports indicate that NETosis is elevated in cases with T1DM or T2DM (15), we compared such activity to that in cases with GDM, using serum cell-free DNA (nucleosomes) as an indicator for NETs release. Although this parameter does not specifically detect NET-derived products in human or murine body fluids, it has been widely used as a surrogate marker for predisposition to NETosis in RA, PE, coagulation disorders, cancer patients, or T1DM and T2DM (24, 29, 31, 33, 41–44). Neutrophils from cases with GDM exhibited the greatest tendency to undergo NETosis, being significantly higher than cases with either T1DM or T2DM (Figure 1C). These results also reaffirmed that pregnancy per se is a pro-NETotic state (Figure 1C) (33). We, next, examined whether other indicators of aspects of neutrophil activity were altered, such as the release of granular enzymes MPO or NE. These were the highest in cases with T1DM (Figures 1D,E).

We were, however, puzzled by the observation that NE release by degranulation was so low in cases with GDM, when compared to normal pregnant controls (Figure 1E). As the EIA assay employed in this analysis detects a complex of NE with its native inhibitor A1AT, we examined the circulating concentrations of the latter, which indicated that these were significantly lower in GDM cases (Figure 1F). As this could lead to an incorrect measurement of NE levels, we accordingly assessed for NE enzymatic activity, which clearly indicated that elastase enzymatic activity was significantly higher in GDM cases than normal pregnant controls (Figure 1G).

Granted that NETosis and NE release was significantly elevated in GDM, our next task was to gain insight into the factors promoting this change in neutrophil activity. A number of reports have indicated that circulatory levels are elevated in cases with GDM and that the potent pro-inflammatory cytokine TNF-α may assist in promoting a pro-diabetic state (45, 46). Our results indicated that circulatory TNF-α levels were indeed elevated in cases with GDM when compared to matching controls (Figure 2C). Since TNF-α is a potent factor for neutrophil activation and NETosis (47), we examined the effect of plasma from GDM cases and normal healthy control pregnancies. This analysis indicated that plasma from GDM cases promoted a greater degree of NETosis on isolated control neutrophils than comparable plasma from healthy pregnant women (Figure 2B). To confirm that TNF-α played a key role in mediating this pro-NETotic effect, we used infliximab, a specific TNF-α antagonist. Such treatment significantly diminished the pro-NETotic effect of GDM plasma on isolated normal neutrophils (Figure 2B).

It has been suggested that the placenta contributes to the pool of elevated TNF-α, which would explain in part the transient pregnancy associated nature of GDM, which usually resolves shortly after delivery (48, 49).

We examined this issue in a trophoblast cell line, BeWo, where we observed that exposure of these cells to hyperglycemic conditions (25 mM glucose) increased the production of TNF-α (Figure 2A). Hence, these data support a placental contribution to the elevated TNF-α levels in GDM.

Furthermore, TNF-α exerted an additive effect on NETosis induced by hyperglycemia in vitro in neutrophils isolated from normal healthy donors (Figure 2D). These findings underscore the potential role of this pro-inflammatory cytokine in neutrophil activation in GDM.

Since an interaction between hyperglycemia and TNF-α on the induction of NETosis was observed in vitro (Figure 2D), we examined whether this phenomenon was also evident in vivo by analyzing samples taken during the OGTT. This test is routinely used to detect GDM in pregnant women, due to their altered response to a HG challenge, as is evident in our study cohort (Figure S2A in Supplementary Material)(50).

In this instance, clear differences in the NETotic capacity of isolated neutrophils were discernible between samples obtained from healthy non-pregnant women, healthy pregnant women, or cases with GDM (Figure 3A), which was most pronounced 60 min after the HG challenge (Figures S2B,C in Supplementary Material). This increase in NETosis was apparent using either Sytox Green to detect exogenous DNA extruded by isolated neutrophils in vitro (Figure 3A, left panel, Figure 3B; Figure S2B in Supplementary Material), fluorescent immunohistochemistry for MPO (Figure 3A, right panel, Figure 3D; Figure S2C in Supplementary Material), as well as by examining for NET products in matching serum samples (Figure S2E in Supplementary Material). As previously observed, this increase in spontaneous NETosis was paralleled by a similar increase in primed pro-NETotic neutrophils, identified by posttranslational histone modifications facilitating chromatin unraveling (citH3⁺) (Figure 3C; Figure S2D in Supplementary Material).

It was also noted that under these conditions, the expression of the NE gene was significantly enhanced in cases with GDM, especially at the 60’ time point of the OGTT assay (Figure 3E).

We, next, examined this interplay between hyperglycemia and TNF-α in a coculture experiment, using BeWo trophoblast-like cells and freshly isolated control neutrophils (Figure 4). Here, we observed that HG conditions (25 mM) promoted neutrophil activation, identified by staining for citH3 (Figures 4A,B), as well as enhanced NET formation (Figures 4A,C) identified by co-staining of extruded nuclear material for NE and citH3. Both neutrophil priming and NET formation were reduced by addition of the TNF-α antagonist, infliximab (Figures 4A–C). In this context, it is of interest that infliximab treatment not only reduced the number of NETs but also had a noticeable influence on their size and dimensions (Figure 4A).

Externally liberated NE has been shown to have a profound effect on neighboring tissues in cancer cells, where its uptake led to the degradation of IRS1, thereby uncoupling the PI3K pathway from the negative regulatory effect of IRS1, thereby leading to uncontrolled growth (36). In a similar manner, exogenous NE was shown to interact with hepatocytes in a murine model for diabetes, where it also enzymatically degraded IRS1 (37), thereby promoting glucose intolerance. For this reason, we examined the influence of exogenous NE on BeWo cells. First, the optimal incubation period of NE with BeWo was determined by monitoring cell viability by means of a MTT assay. Influence of NE on BeWo indicated that akin to the report on cancer cells (36), exogenous NE leads to a significant degradation of IRS1 protein (Figure 4D). This effect of NE is also evident in cultures treated with insulin (Figure 4D). The decrease in IRS1 protein levels was associated with a concomitant decrease in glucose transporter type 4 gene (GLUT4), resulting in diminished glucose uptake (Figure 4E). Once again, this effect was also present in insulin-treated cultures, albeit to a lower level than cultures without insulin (Figure 4E).

Our next task was to ascertain whether such an interaction occurs in the placenta of pregnancies affected by GDM. A histological analysis indicated that there were numerous neutrophil infiltrates in the villous tissue in GDM placentae (Figure 5A), which was confirmed by fluorescent immunohistochemistry for MPO and citH3 (Figures 5B–D), suggestive of primed and NETing neutrophils (Figures 5A–D).

We also determined that GDM placentae contained elevated numbers of NE expressing cells at the protein and de novo mRNA synthesis (Figures 5F–H). Akin to our observations in BeWo cells, we noted that the elevated NE was associated with a decreased presence of IRS1, as determined both by western blotting (Figure 5E) and fluorescent immunohistochemistry (Figure 5F).