This study has been designed and conducted according to the principles of the Declaration of Helsinki and approved by the Ethics Committee of the Università Cattolica del Sacro Cuore of Rome, Italy. Written informed consent was obtained from all recruited individuals.

All women enrolled in this study were selected among those referring to the High Risk Pregnancy Unit and to the delivery suite of the Gemelli Hospital of Rome. For the identification of cases of PE we referred to the definition of the International Society for the Study of Hypertension in Pregnancy (ISSHP) (7). Control patients were recruited among those attending the Gemelli Hospital as outpatients for pregnancy routine medical examinations. Women with diabetes, obesity (BMI >30), pre-existing hypertension, autoimmune or infectious diseases or fetal abnormalities were excluded from control group. All cases and controls were matched for gestational age.

3 ml of venous blood was collected by venipuncture from the antecubital fossa in a tube with EDTA. Each sample was centrifuged twice at 3000 g for 30 minutes at 4°C, to obtain plasma, and then frozen in 500 μl aliquots at -80°C until use.

Plasma samples (500µl), previously stored at -80°C, were allowed to attain room temperature. Each sample was diluted (1:1 vol/vol) with PBS and spun at 120,000 g for 90 minutes at 4°C in an Optima XPN ultracentrifuge (Beckman Coulter). After the ultracentrifugation step, the supernatant was discarded and the pellet, which comprised of extracellular vesicles of different sizes, was washed with PBS and spun one more time at 120,000 g for 90 minutes at 4C°. The supernatant was then discarded and the pelleted EVs were then re-suspended with 100 µl of PBS. The protein concentration of STBEVs was determined by Bradford assay, prior to storage at −80°C.

For PLAP expression analysis, STBEVs samples collected from sera from 4 preeclamptic and 4 control patients were lysed on ice for 30 minutes in SDS-PAGE sample buffer containing protease inhibitor cocktail (Roche Diagnostics,Basel, Switzerland). Protein content was normalized by loading 20 μg of protein for each sample. Samples were boiled and centrifuged at 13,000g for 10 minutes prior to separation by SDS/PAGE (Invitrogen) and semi-dry transfer to PVDF membrane (Biorad, Hercules, CA, USA). Non-specific binding was blocked with TBS-T (20 mM Tris/HCl, 137 mM NaCl, 0.1%Tween-20, pH 7.6) containing 5% of milk (Santa Cruz Biotechnology Inc., Dallas, Texas, USA) for 1 hour at R/T. Membranes were incubated at 4°C overnight with the NDOG2 antibody (1µl/ml) syncytiotrophoblast-specific mouse monoclonal antibody that recognizes placental alkaline phosphatase (PLAP) (8) in Tris-buffered saline and 0.05% Tween 20 (TBS-T) containing 1% BSA. Membranes were then washed in TBS-T, before incubation with the appropriate (rabbit or mouse) horseradish peroxidase-conjugated secondary antibody (1:4000; Dako, Glostrup, Denmark) for 1 hour at R/T. The antibody used was diluted in blocking buffer. After washing, blots were treated with an enhanced chemiluminescence system (PierceTM, Thermo Fischer Scientific, Waltham, MA USA) and exposed to Hyperfilm ECL (GE Healthcare Life Sciences, Cleveland, Ohio, USA). Densitometric analysis of Western blot was carried out using NineAlliance software (Uvitec Alliance, Cambridge, UK).

Analysis of STBEVs was carried out by multi-color flow cytometry, using a CytoFLEX S cytometer (Beckman Coulter) equipped with violet laser (405 nm) excitation sources. This instrument is able to collect SSC (side scatter) off the blue laser (BSSC) and the violet laser (VSSC). The flow cytometer was calibrated using the Megamix-Plus FSC beads emitting FITC of different sizes (100, 300, 500, and 900 nm), as described elsewhere (9). STBEVs number was measured using the cell-counting feature of the instrument that relies on a calibrated peristaltic pump for sample delivery. To distinguish intact vesicles from cell debris, STBEVs were stained with CellTrace™ Calcein Violet (ThermoFisher Scientific, Waltham, Massachusetts, USA), according to manufacturer instructions, as previously described (9).

Prior to running STBEV samples, filtered PBS was also analyzed in triplicate for 2 min to assess the level of background contaminating events (about 100 events/second). To confirm placental origin, STBEVs were stained with anti-Placental Alkaline Phosphatase (PLAP) NDOG2 mouse monoclonal antibody commercially conjugated to Phycoerythrin (PE) (Biolegend UK Ltd., Cambridge, UK) or its PE-conjugated IgG1 control (Biolegend). PLAP is exclusively expressed in placental tissues and commonly used to distinguish vesicles released by the placenta from those coming from other cell types. PLAP positive vesicles were normalized to calcein positive events. Vesicles collected from culture medium of a colorectal cancer cell line were used as a negative control for PLAP positivity. The presence of HLA-DR on STBEVs was investigated by the binding of fluorescein isothiocyanate (FITC)-conjugated mouse monoclonal anti-HLA-DR antibody (clone L243-Abcam) or IgG1 control (Abcam) antibody. Prior to use, all antibodies were filtered through Nanosep 0.2 μm centrifugal devices (Pall Life Sciences) to minimalize interference by background particles. All antibodies were also titrated to ensure their use at the optimum concentration. Fluorochrome compensation for multicolor STBEV flow cytometry was set-up using BD CompBeads (BD Biosciences) labelled with fluorescence conjugated antibodies.

All samples were then blocked with 0.2 µm filtered Fc receptor blocker (10 µL; Miltenyi) for 10 minutes at 4°C before the addition of either filtered PBS (unlabeled control), fluorescence conjugated IgG control (IgG1-PE); or fluorescence conjugated antibody (NDOG2-PE or HLADR-FITC). Tubes were incubated for 1 hours at 4°C in the dark. Stained samples were then made up to 500 µL with filtered PBS and acquired immediately on the flow cytometer (Beckman Coulter). Gates were firstly set so that ≤1% of STBEVs stained positive in the appropriate negative controls. Data were acquired and analyzed by the CytExpert 2.2™ software (version 2.2, CytoFLEX S, Beckman Coulter, Milano, Italy).

Clinical characteristics were analyzed using unpaired t-test or Chi-square, according to type of variables. The results of the experiments, expressed as the mean ± standard deviation (SD), were analyzed using unpaired t-test performed with Prism software version 9.0. For all analyses, p < 0.05 was considered significant.

Peripheral venous blood was obtained by venous puncture from 22 women with diagnosis of PE and 22 normal pregnant (NP) women with uncomplicated pregnancies, matched for gestational age. Clinical characteristics of PE and NP subjects are shown in Table 1 . PE women were more frequently nulliparous (p < 0.5) and displayed significantly lower gestational ages at delivery, neonatal birth weights (p < 0.0001) and birth weight percentile (p < 0.0001) compared to NP. As expected, women with PE were characterized by higher blood pressures (p < 0.0001) and detectable proteinuria (p < 0.01) compared to NP women. No significant differences were found in terms of maternal age between PE and control groups. 52% of women with PE were primiparous versus 31% of controls.

Western blot analysis of STBEVs obtained by plasma ultracentrifugation, performed to assess the presence in maternal circulation of detectable placental-derived vesicles, revealed a clear protein band corresponding to placental alkaline phosphatase (PLAP), a specific marker of syncytiotrophoblast. These results confirmed the presence of significant quantity of STBEVs in the plasma of both PE and NP women ( Figures 1A, B ). Vesicles collected from culture medium of a colorectal cancer cell line, and used as a negative control for PLAP positivity, showed, as expected, no positivity for PLAP (data not shown). Flow cytometric analysis, performed to quantify serum STBEVs content, showed that, according to available literature (10–12), levels of circulating STBEVs were significantly higher in women with PE compared to normal pregnant women ( Figures 1C–G ), although STBEVs levels showed a high standard deviation (SD) and a skewed distribution among the study groups. Interestingly, STBEVs were detectable in the maternal circulation from the early first weeks of gestation ( Figure 1D ). No significant differences were found in terms of calcein positive events or protein content between PE and NP women in vesicles samples collected by plasma ultracentrifugation (data not shown).

Flow cytometric analysis revealed that HLA-DR positive circulating STBEVs can be detected in PE but not in NP women (p<0.01) ( Figures 2A–D ). In particular, we found a subset of 14 of 22 PE (64% of all PE women analyzed) with aberrant expression of HLA-DR on circulating STBEVs. None of NP women showed detectable positivity for HLA-DR in circulating STBEVs. These data were consistent with the results previously observed on STBEVs collected by dual placental perfusion (3). Interestingly, HL-DR was detectable on STBEVs at all gestational ages analyzed (from a very rare case of PE at 15 weeks until term of gestation) ( Figure 2B ).

When comparing clinical characteristic of PE women with STBEVs carrying HLA-DR with those with HLA-DR negative STBEVs, no significant differences were found ( Table 2 ).