Human placental tissue samples [normal, n = 10, gestational age at delivery (weeks): 39.3 ± 0.7, and PE, n = 10, gestational age at delivery (weeks): 38.5 ± 0.9] were obtained between 29 and 40 weeks of gestation from Jan to Dec of 2019 and provided by Tongde Hospital of Zhejiang Province (Hangzhou, China), following a protocol approved by the Ethics Committee of Tongde Hospital of Zhejiang Province. Patients with PE were defined as having systolic and diastolic blood pressure > 140 and 90 mm Hg, respectively, measured at least 6 h apart plus proteinuria ≥ 300 mg/24 h or > 1+ by dipstick test. Clinical states of known pathogenesis which could possibly interfere with values of studied parameters were excluded. Exclusion criteria were also uterine contractions and premature rupture of amniotic membranes. To get similar clinical condition with PE, control group with similar gestational age were collected. Written informed consent was obtained for each participant.

Dihydroethidium (DHE, O₂·) was determined using a DHE assay kit (Beyotime, China) as described previously (Huang et al., 2013). Placental explants were washed with PBS, incubated in fresh culture medium containing 2 μM DHE for 30 min at 37°C and washed three times. Fluorescence intensity was measured with a fluorescence microscope, with the excitation wavelength of 535 nm.

Dry materials of the ingredients in proportions as listed here were pulverized. One dose of TGD for human is: Gastrodiae Rhizoma (Tianma) 9 g, Uncaria Ramulus Cum Uncis (Gouteng) 12 g, Haliotidis Concha (Shijueming) 18 g, Gardeniae Fructus (Zhizi) 9 g, Scutellariae Radix (Huangqin) 9 g, Cyathulae Radix (Chuanniuxi) 12 g, Eucommiae Cortex (Duzhong) 9 g, Leonuri Herba (Yimucao) 9 g, Taxilli Herba (Sangjisheng) 9 g, Polygoni Multiflori Caulis (Shouwuteng) 9 g, Poria (Fuling) 9 g (The State Pharmacopoeia Commission of P. R. China, 2015). After pulverizing, the powder was steeped with threefold tap water (v/w) for 0.5 h, boiled for 1 h, and then filtered. The residue was extracted with tap water and the filtrate was concentrated by rotary evaporation under vacuum in a 60°C water bath. The concentrated extract was frozen in liquid nitrogen and finally subjected to lyophilisation under vacuum. Final yield was powdered and then stored at −20°C. All medical materials were purchased from Zhejiang Academy of Traditional Chinese Medicine.

The human first-trimester extravillous trophoblast TEV-1 cell line was obtained from the American Type Culture Collection (ATCC, Manassas, VA) and was incubated in Dulbecco’s 1:1 modified Eagle’s medium/Ham’s F12 culture medium (Life Technologies) supplemented with 10% (v/v) fetal bovine serum (FBS, Life Technologies, Inc., Grand Island, NY) at 37°C in an atmosphere of 5% CO₂.

Antibodies for nNOS (mouse, sc-5302), iNOS (mouse, sc-7271) and cGMP (mouse, sc-21727) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA), eNOS antibody (rabbit, AF6792) and αTubulin antibody (rabbit, AF0001) were from Beyotime Biotechnology (Shanghai, China), and antibodies for ICAM1 (rabbit, bs-4615R) and NF-κB (mouse, bsm-33059M) were from Bioss (Beijing, China). H₂O₂ was obtained from Beyotime Biotechnology (Shanghai, China) and L-NAME was from Selleck (Shanghai, China). ShRNA sequences used are as follows: shRNA-nNOS (CAG​AGA​ATA​GTT​CAC​ATC​TAT), shRNA-iNOS (ATC​GAA​TTT​GTC​AAC​CAA​TAT), shRNA-eNOS (CAG​GAA​GAA​GAC​CTT​TAA​AGA), and a scrambled shRNA sequence (TTT​GTA​CTA​CAC​AAA​AGT​ACT​G) was used as control.

In order to confirm the appropriate concentration and time-course of TGD administration, we treated TEV-1 cells with three different concentrations, including L (25 mg/ml), M (50 mg/ml) and H (100 mg/ml) (Liu et al., 2015), and with different time periods, including 12, 24 and 48 h. Cell Counting Kit-8 (CCK-8) assay was employed to evaluate cell proliferation and the concentration and treatment period of TGD that had no significant effect on TEV-1 cells proliferation was chosen for further experiments.

CCK-8 assay was performed according to the manufacturer’s protocol (Yeasen, Shanghai, China) as described previously (Jian et al., 2021). After treatment, TEV-1 cells were incubated with 10 μl CCK-8 reagent for 2 h at 37°C and the absorbance was measured at a wavelength of 450 nm with a micro-plate reader.

TEV-1 cells were seeded into the 96-well plate at the density of 5 × 10³ cells per well, and cultured at 37°C for 24 h. After the treatment with different concentrations of TGD for 24 h, TEV-1 cells were exposed to different concentrations of H₂O₂ (150 μM, 300 μM or 500 μM) for 3 h in the presence or absence of TGD. Six replicated wells were set for each group.

Intracellular ROS levels were measured by oxidation-sensitive fluorescent probe 2′,7′-dichlorofluorescein-diacetate (DCFH-DA) following the instructions (Beyotime Biotechnology, Shanghai, China). Briefly, TEV-1 cells were cultured in 96-well plates and were collected after treatment. Then, the TEV-1 cells or placental tissue explants were washed with PBS, and were subsequently incubated in 10 μM DCFH-DA for 20 min at 37°C. Fluorescence was quantified by a micro-plate reader with the excitation wavelength of 488 nm and the emission wavelength of 525 nm.

The levels of intracellular NO were measured using the Nitric Oxide Assay Kit (Applygen Tech., Beijing, China) as previously reported (Ji et al., 2013). Briefly, TEV-1 cells were seeded in 60 mm plates and incubated overnight. Then TEV-1 cells were transfected with control shRNA, nNOS shRNA, iNOS shRNA or eNOS shRNA, and treated with 500 μM H₂O₂ for 3 h or with 100 μM L-NAME for 12 h, followed by the incubation with TGD for 24 h. After incubation, Griess reagent was added and the resultant color development was quantified by reading the absorbance at 550 nm. For measurements in placental samples, placental explants were cultured in 12-well plates for 48 h, then the medium was collected and the Griess reagent was added. The resultant color development was quantified by reading the absorbance at 550 nm.

The activity for SOD was determined using the xanthine oxidase assay kits (Nanjing Jiancheng Bioengineering Institute, China) as described previously (Mao et al., 20122012). Briefly, when the treatment was done, TEV-1 cells were washed thoroughly, collected, and lysed on ice with an ultrasound disruptor. Thereafter, the lysed cells were centrifuged at 10,000 g for 10 min, and the total protein concentration of the supernatant was tested using a bicinchoninic acid (BCA) kit (Beyotime Biotechnology, Shanghai, China). The xanthine/xanthine oxidase system was used to produce superoxide anion (O²⁻), which react with 2-(4-iodophenyl)-3-(4-nitrophenol-5-phenlyltetrazolium chloride) to generate a red formazan dye, and the absorbance was measured at a wavelength of 450 nm.

CAT activity was measured using a kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) according to the manufacturer’s instructions. Briefly, when the treatment was done, TEV-1 cells were washed thoroughly, collected, and lysed on ice with an ultrasound disruptor. Thereafter, the lysed cells were centrifuged at 10,000 g for 10 min, and the total protein concentration of the supernatant was tested using a bicinchoninic acid (BCA) kit (Beyotime Biotechnology, Shanghai, China). H₂O₂ is a substrate of a variety of enzymes in cells, including CAT. The CAT activity was thereby determined by analyzing the decomposition rate of H₂O₂ caused by CAT and is calculated as units per grams of protein (U/g protein).

The MDA activity was measured using a spectrophotometric diagnostic detection kit (Nanjing Jiancheng Biotechnology Institute, Nanjing, China) as described previously (Li et al., 2014). Briefly, when the treatment was done, TEV-1 cells were washed thoroughly, collected, and lysed on ice with an ultrasound disruptor. Thereafter, the lysed cells were centrifuged at 10,000 g for 10 min, and the total protein concentration of the supernatant was tested using a bicinchoninic acid (BCA) kit (Beyotime Biotechnology, Shanghai, China). MDA was allowed to react with thiobarbituric acid that yielded red-coloured products that were spectrophotometrically quantified by measuring the maximum absorption peaks at 532 nm.

Determination of GSH-PX activities was carried out by kits according to the manufacturer’s instructions (Nanjing Jiancheng Biotechnology Institute, Nanjing, China) as described previously (Lin et al., 2015). Briefly, when the treatment was done, TEV-1 cells were washed thoroughly, collected, and lysed on ice with an ultrasound disruptor. Thereafter, the lysed cells were centrifuged at 10,000 g for 10 min, and the total protein concentration of the supernatant was tested using a bicinchoninic acid (BCA) kit (Beyotime Biotechnology, Shanghai, China). The results were calculated in units of enzymatic activity per milligram of protein (U/mg protein).

The activities of TNOS were assayed using a TNOS assay kit (Nanjing Jiancheng Biotechnology Institute, Nanjing, China) as described previously (Qiu et al., 2018), based on NOS catalysis that converts amino acid L-arginine into citrulline and generates NO. Briefly, when the treatment was done, TEV-1 cells were washed thoroughly, collected, and lysed on ice with an ultrasound disruptor. Thereafter, the lysed cells were centrifuged at 10,000 g for 10 min, and the total protein concentration of the supernatant was tested using a bicinchoninic acid (BCA) kit (Beyotime Biotechnology, Shanghai, China).

Total RNA was extracted from TEV-1 cells according to the instructions of the Trizol reagent according to the manufacturer’s instructions (Yeasen, Shanghai, China). 5 μg total RNA in a volume of 20 μl was reversely transcribed to generate cDNA using SuperScript III reagent (Takara Biotechnology, Dalian, China). And then 2 μl of cDNA were used for per quantitative real-time PCR amplification, which was performed on an ABI 7500 Real-Time PCR System (Applied Biosystems, Waltham, MA) according to the qRT-PCR SYBR Green method (Yeasen, Shanghai, China) with the following primers: nNOS, F-5′-GGAATCCAGGTGGACAGAGA-3′, R-5′- GGG​CAG​AGG​TTT​GTG​TGA​CT-3’; iNOS, F-5′-CAGCGGGATGACTTTCCAAG-3′, R-5′- CAG​CGG​GAT​GAC​TTT​CCA​AG-3’; eNOS, F-5′-GGACTTCATCAACCAGTAC-3′, R-5′- GAT​GTA​GGT​GAA​CAT​TTC​C-3’; ICAM1, F-5′-GGCATTGTTCTCTAATGTCTCCG-3′, R-5′- CCG​CTC​AGA​AGA​ACC​ACC​TTG​G-3’; sFLT1, F-5′-ACAATCAGAGGTGAGCACTGCAA-3′, R-5′- TCC​GAG​CCT​GAA​AGT​TAG​CAA-3’; αTubulin, F-5′-ATGCCAAGTGACAAGACCATTG-3′, R-5′- ACT​TCT​TGC​CAT​AAT​CAA​CTG​AGA-3’. The CT value was recorded and the relative expression levels of target genes were normalized to αTubulin levels, respectively, to the analyzed using the 2^−ΔΔCt method. Each experiment was performed in triplicate.

Western blots were performed as described previously (Jin et al., 2021). Briefly, TEV-1 cells were harvested and lysed using NP-40 buffer in the presence of protease inhibitors cocktail (Beyotime Biotechnology, Shanghai, China). Protein concentrations were measured by a bicinchoninic acid (BCA) protein assay kit (Beyotime Biotechnology, Shanghai, China). Then protein was subjected to SDS-PAGE Gel (50 μg protein per lane) and was transferred to polyvinylidene difluoride (PVDF) membranes. Membranes were firstly blocked with 5% non-fat milk for 1 h at room temperature followed by incubation with the corresponding primary antibodies overnight at 4°C. Then membranes were incubated with the secondary antibodies linked to horseradish peroxidase for 1 h at room temperature, and signals were subsequently developed using the Enhanced Chemiluminescence System.

Adult male and female C57BL/6 mice weighing 18–20 g were obtained from Cyagen (Suzhou, China), and were provided free access to standard mouse chow and tap water and were maintained under controlled conditions of temperature (23°C), humidity (40–60%), and lighting (12 h/12 h light/dark cycle) for 1 week before the experiments. Vaginal smears were obtained daily from female mice until the vaginal cytologic findings were indicative of estrus. The female mice were mated with male mice at a ratio of 2:1 and assessed for vaginal plugs the following morning, which was designated gestational day 0 (GD 0). On GD 0.5, the female mice were assigned to 4 groups (n = 8 each). From gestational day 7.5 (GD 7.5) to GD 18.5, L-NAME (60 mg/kg/d; Selleck, Shanghai, China) or normal saline (similar dose) was administered subcutaneously, and from GD 8.5 to GD 18.5, TGD (dissolved in pure water) at 20 g/kg/d (low dose) or 40 g/kg/d (high dose) or pure water was intragastrically administered (Deng et al., 2020).

Systolic blood pressures (SBPs) were measured before grouping and on GDs 6.5, 12.5 and 18.5 via the indirect method of tail cuff occlusion in conscious animals using a CODA monitor system (Kent Scientific, Torrington, CT) as described previously (Singh et al., 2011). Before each measurement was taken, each mouse was exposed to 37°C for 5 min, and measurements were taken three times for averaging purposes.

Twenty-four-hour urine was collected with metabolic cages on GDs 6.5 and 18.5. The urine outputs were recorded and the concentration of protein was examined by using a bicinchoninic acid (BCA) protein assay kit (Beyotime, China) according to the manufacturer’s instructions and was presented in absorbance units (AU). Briefly, 10 µl of the samples and standards (bovine serum albumin) were used, and afterwards, 200 µl of the working solution (bicinchoninic acid and copper sulphate, 49:1 ratio, respectively) were added. Then, samples were incubated at 37°C for 30 min. After cooling the samples at room temperature, absorbance was measured at 562 nm. Finally, the 24-h urine protein concentration was calculated and was normalized by urine volume.

Pregnancy was terminated on GD 18.5 via cesarean delivery after pregnant dams were killed by cervical dislocation under isoflurane. The placentas and delivered pups from the horns of the uterus were examined for weight, length and number, and the resorptions were inspected as well. The isolated placentas were washed with an ice-cold 0.9% saline solution, fixed in 4% paraformaldehyde (PFA) and embedded in paraffin. Four-micrometer paraffin sections were prepared, stained with hematoxylin-eosin and observed under a light microscope (Nikon, Tokyo, Japan).

Data are presented as means ± S.D. and were analyzed by an unpaired two-tailed t test, one-way ANOVA and two-way ANOVA (SPSS 13.0J software; SPSS, Inc., Chicago, IL). Statistical significance was assessed at p < 0.05 and p < 0.01.

To investigate the role of TGD in PE treatment, we cultured TEV-1 cells, which were derived from the first-trimester normal extravillous trophoblast cells of a healthy woman placenta, and treated with TGD. In order to determine the appropriate concentration and treatment time for further experiments, we performed CCK-8 assays to first confirm the effect of TGD on TEV-1 cells growth. The results showed that treatment with TGD of either medium dose (M, 50 mg/ml) or high dose (H, 100 mg/ml) significantly promoted TEV-1 cells proliferation at both 12 and 24 h, and high dose TGD-treatment (H, 100 mg/ml) also caused a significant increase in cell proliferation at 48 h, whereas the low dose groups (L, 25 mg/ml) at 12, 24 or 48 h did not present obvious alterations in TEV-1 cell proliferation (Figure 2A). Thus, we chose TGD treatment of low dose (25 mg/ml) for 12 h or 24 h for further experiments.

To next investigate the potential effect of TGD on oxidative stress, H₂O₂ was utilized to stimulate ROS generation in TEV-1 cells to mimic the oxidative stress in PE trophoblasts, as superoxide anion generated endogenously or exogenously can rapidly be converted to hydrogen peroxide such as H₂O₂ (Veal et al., 2007), and previous studies showed a significant elevation of H₂O₂ levels in the bloodstream as well as the placentas of women with PE (Emberton et al., 2003; Steiner and Raghow, 2003; Zhou et al., 2012; Cross et al., 2015). Results from CCK-8 assays revealed that only H₂O₂ treatment at 500 μM for 6 h significantly impeded TEV-1 cells proliferation (Figure 2B), while H₂O₂ treatment at 500 μM for both 3 and 6 h significantly caused ROS production (Figure 2C). Thus, H₂O₂ treatment at 500 μM at 3 h was chosen for further investigations.

Although the H₂O₂ treated (500 μM, 3 h) group showed significantly increased ROS production and decreased NO levels (Figures 2D,E), treatment with TGD for 24 h significantly suppressed the ROS levels in the presence or absence of H₂O₂ stimulation (Figure 2D), while TGD treatment for 24 h not only significantly elevated NO production but also effectively restored the H₂O₂-inhibited NO levels (Figure 2E), suggesting the role of TGD in relief from oxidative and nitrosative stresses in trophoblasts. Consistently, in PE placental tissues, the mRNA expression of intercellular adhesion molecules, such as intercellular adhesion molecule 1 (ICAM1) (Marzolla et al., 2017), and anti-angiogenic factors, such as soluble Fms-Like Tyrosine kinase-1 (sFLT1) (Maynard et al., 2003), was significantly up-regulated (Figure 2F), whereas TGD treatment not only significantly enhanced the content levels of cGMP, the second messenger cyclic guanosine-3′,5′-monophosphate (cGMP) and a key downstream effector of NO (Kastrati et al., 2010), but also significantly suppressed the mRNA and protein expression of ICAM1, decreased sFLT1 mRNA expression and inhibited the protein levels of oxidative stresses-regulated transcription factors such as NFκB (Figures 2G–I), indicating that TGD exerts antioxidant and anti-inflammatory effects.

Given that oxidative stress can also result from a lack of antioxidants, including the enzymatic antioxidants, the major category of antioxidants, such as catalase (CAT), SOD and GSH-PX (Espinosa-Diez et al., 2015), we further examined the effect of TGD on the activities of these antioxidants. Unexpectedly, the activities of CAT, SOD or GSH-PX showed no change in response to TGD treatment, nor did the MDA activity, a molecule involved in oxidative stress pathway (Figure 2J) (Khoubnasabjafari et al., 2015). Thus, TGD prevents ROS generation and promotes NO production.

NO is synthesized from L-arginine under the action of NOS, which catalyze L-arginine oxidation into NO and L-citrulline (Knowles and Moncada, 1994; Förstermann and Sessa, 2012). To explore the mechanisms underlying TGD in regulation of NO production, we examined the effect of TGD on TNOS enzymatic activity using a commercial kit. Our results exhibited that H₂O₂ treatment (500 μM, 3 h) significantly suppressed TNOS activity, which was consistent with a previous report (Kotsonis et al., 1999), while TGD treatment not only significantly up-regulated TNOS activity but also effectively restored the H₂O₂-suppressed TNOS activity (Figure 3A), indicating TGD regulates TNOS activity. Consistently, treatment with L-NAME, an inhibitor of TNOS activity (Pfeiffer et al., 1996), significantly lowered the TGD-induced NO levels, compared with the TGD-treatment group (Figure 3B). Thus, TGD promotes NO generation by up-regulation of TNOS activity.

Given that there are at least three NOS proteins encoded by distinct genes in human placenta, including nNOS, immunological (inducible) nitric oxide synthase (iNOS), and eNOS (Di Iulio et al., 1999; Yang et al., 2005), we next want to know whether all the three NOS isoforms are involved in TGD-regulated NO generation. To this end, we first examined the corresponding expression of the three NOS in response to TGD treatment, respectively. Unexpectedly but intriguingly, TGD significantly up-regulated the mRNA and protein levels of nNOS and eNOS, while iNOS expression remained unchanged in the presence of TGD (Figures 3C,D). Moreover, H₂O₂ treatment (500 μM, 3 h) suppressed the mRNA expression of all the three NOS isoforms, among which, the expression of nNOS and eNOS could be effectively attenuated by TGD treatment for 24 h (Figure 3E), indicating nNOS and eNOS may participate in TGD-mediated NO production. On the other hand, we constructed plasmids expressing NOS-shRNAs which knocked down the mRNA expression of nNOS, iNOS or eNOS by as much as 40–60% (Figure 3F), and observed the alterations in NO production in NOS shRNAs-transfected TEV-1 cells, respectively. Consistently, knockdown of nNOS or eNOS but not iNOS significantly down-regulated the TGD-induced NO levels (Figure 3G). Thus, TGD promotes NO generation by up-regulation of the expression of nNOS and eNOS.

To verify the therapeutic effect of TGD on PE in vivo, we next took advantage of the L-NAME-induced PE mouse model as described previously (Zhao et al., 2018). Schematization of the procedures and treatments was shown in Figure 4A. Results revealed that the mean systolic blood pressure (SBP) on GD 6.5 were 110 ± 1.6 mm Hg in the control group without pregnant and 112 ± 1.6 mm Hg in the group of pregnant mice awaiting L-NAME treatment (Figure 4B). Once L-NAME was administered, the mean SBP of the pregnant mice in the L-NAME group was increased significantly on GD 12.5 and maintained until GD 18.5, compared with the control group (Pregnancy, vehicle) (Figures 4C,D), while administration of TGD at both low dose and high dose significantly lowered the SBP on GD 12.5 and GD 18.5 , compared with the L-NAME group (Figures 4C,D), suggesting the hypotensive effect of TGD during pregnancy. Although the 24-h urinary protein concentration did not show significant change between the control group (Pregnancy, vehicle) and the L-NAME group on GD 12.5 (Figure 4E), administration of L-NAME significantly increased the 24-h urinary protein concentration on GD 18.5 (Figure 4F), which was effectively restored by administration of TGD at a high dose (Figure 4E). Of note, neither the mean SBP nor the 24-h urinary protein concentration in the control group (Pregnancy, vehicle) was significantly altered by TGD administration (Figures 4C–F).

Compared with the control group (vehicle), the number of resorptions in the L-NAME group was increased, which was down-regulated by TGD administration at either a low dosage or a high dosage (Figure 5A), while no malformations were observed in the fetuses in all groups (Figure 5B). In addition, the mean weight of the fetuses, the crown-rump length of the fetuses and weight of the placentas in the L-NAME group were significantly decreased compared with the control group (vehicle), which were partially up-regulated by TGD administration of both dosages (Figures 5C–F). Although the diameters of the placentas did not present significant differences (Figure 5G), morphologic changes in the placenta could be observed under a light microscope (Figures 5H–J). Placental villi formed a complicated labyrinth-like structure with maternal blood cells, which occupied the spaces of the mesh and contributed to the exchange of the material and nutrition between the maternal and fetal blood. On GD 18.5, normal histologic structures were observed in the control group (vehicle), whereas increased syncytial knots were noted in the L-NAME administration group (Figures 5H–K), resulting in the consequently decreased blood flow in the placenta. However, noticeably, the number of syncytial knots was reduced when TGD administration particularly at the high dosage was applied in the L-NAME group, compared with L-NAME administration alone (Figures 5H–K). Additionally, we also examined the NO levels in mouse placentas and our results exhibited that, although L-NAME significantly suppressed NO levels, TGD administration effectively retrieved that to a comparable level to the vehicle control group (Figure 5L), which was in accordance with the in vitro data (Figure 3B). In summary, TGD restores the pregnancy outcome in a PE mouse model.