The 188 patients recruited were those who received ART treatment between April, 2012 and November, 2016 at Luoyang Central Hospital Affiliated to Zhengzhou University (Luoyang, China). All protocols associated with humans in the present study were approved by the Review Board of Luoyang Central Hospital Affiliated to Zhengzhou University. All participants provided permission to cooperate to undertake the relevant study and provided written informed consent. The study included 188 cycles. The inclusion criteria were as follows: Age, 20-38 years, embryo freezing storage period between 30 and 1,080 days, vitrification was applied as an embryo cryopreservation method. Patients with a history of chronic hypertension (and/or used antihypertensive medication prior to pregnancy), multifetal gestation, or metabolic dysfunctions were excluded from the study. According to standard criteria (26), PE was defined as gestational hypertension with proteinuria.

The study was performed under the approval of the Animal Ethics Committee of Luoyang Central Hospital Affiliated to Zhengzhou University. The PTN^−/− mice were generated by deleting exons 2-4 on a background of 129/OlaxC57BL/6 J as described in two previous studies (28,29). The animals (614 female mice, 8-10 weeks old, weighing 18-25 g) obtained from Nanjing Biomedical Research Institute (Nanjing, China) were divided into [WT (304 mice) group and PTN^−/− (310 mice) group] for the preparation of false pregnancy. The animals had free access to food and water and were housed at 22°C (60-70% humidity, 12 h light/12 dark). The average body weight of the mice in each group was between 20 and 25 g (8-10 weeks of age). The genotype of PTN^−/− mice was identified using a polymerase chain reaction method as previously described (30). The primers used were as follows: Forward, 5′-GATTGAACAAGATGGATTGC-3′ and reverse, 5′-CAT TTAGGCAAACAGGAAGGACG-3′. The genomic DNA extracted from tails of PTN^−/− (a total of 310 mice) and WT mice (a total of 304 mice) was used as the template. The temperature protocols was set as: 94°C, 5 min; 35 cycles of 94°C, 30 sec, 63°C, 30 sec, 72°C 1 min; final extension at 72°C, 10 min.

An intraperitoneal injection of permanent- magnet synchronous generator (PMSG) (5 IU/mice; Ningbo Sansheng Pharmaceutical Co., Ltd., Ningbo, China) was administered to female mice (75 mice, 8 weeks old, weighing 18 g). After 48 h, human chorionic gonadotropin was then injected (5 IU/mouse; Ningbo Sansheng Pharmaceutical Co., Ltd.). The eggs were collected 14 h later. In brief, the female mice (weight, 18 g) were sacrificed by cervical dislocation and following disinfection, the fallopian tubes were rapidly isolated and washed in PBS working buffer. The abdomen of the fallopian tube was then opened with ophthalmic forceps under a stereomicroscope. The egg granule cell complex was collected and washed in PBS and HTF medium (Quinn’s; SAGE-In vitro Fertilization, Inc., Trumbull, CT, USA). The sperm from the cauda epididymis of male mice (25 mice, 8 weeks old) were incubated in HTF medium covered with paraffin oil (Sigma; EMD Millipore, Billerica, MA, USA) at 37°C for 1 h (5% CO₂ atmosphere). Following this treatment, the sperm (final sperm concentration, 2.0×10⁶/ml) were co-cultured with eggs droplets (30-40/droplets) for 4-6 h for in vitro fertilization (fertilization rate, 74.36%). The fertilized eggs in cleavage medium covered with paraffin oil (Sigma; EMD Millipore) were maintained for 36-48 h in order to acquire four-cell stage embryos. Blastula medium (BM; SAGE-In vitro Fertilization, Inc.) was used to culture four-cell embryos for obtaining blastocysts.

In brief, frozen liquid containing vitrification solution (VS) solution and equilibration solution (ES) (both from Kitazato, Tokyo, Japan) were prepared 30 min in advance. The blastocysts were maintained in ES for 10 min and in VS for 30 sec. Finally, the embryos were transferred onto the top of Cryotop (Kitazato) with a minimal volume of embryo solution attached, which were immediately stored in liquid nitrogen. The contact time of the embryo with the VS liquid was between 30 and 60 sec. For thawing, the blastocysts maintained in Cryotop were collected and directly immersed in thawing solution for 60 sec, and then in diluent solution for 60 sec, in washing solution (WS)1 for 3 min, and in WS2 for 3 min. The blastocysts were then transferred into BM covered with mineral oil for 1.5-2.5 h.

The male mice (8 weeks old, weighing 20-22 g, Nanjing Biomedical Research Institute) were vascularized under the anesthesia status. The mice were anesthetized by the intravenous administration of sodium pentobarbital solution (2 mg/ml; Solarbio, Beijing, China) (30-50 mg/kg body weight). In brief, following anesthesia, the mice were disinfected with 70% ethanol. The vas deferens of the mice was ‘picked up’ with tweezers and was ligated. The vas deferens was then cut off in the middle of the two ligatures. The healed vasectomized male mice were mated with estrous female mice from the WT group and PTN^−/− group as already described. The following morning, the female mice with copulatory plugs were selected as recipients for pseudopregnancy for 12 h. The frozen-thawed blastocysts were then transferred into the recipients at 2.5 days of pseudopregnancy.

Following the blastocyst transfer, the pregnant WT and PTN^−/− mice were injected intraperitoneally with either corn oil taxsolution (3 mg/40 g body weight; Sigma; EMD Millipore) or an equal volume of corn oil vehicle solution on day 10 of pregnancy. The tax was injected for 5 days consecutively. This dose regimen has been reported previously (31). Four groups were established in terms of the genotype and treatment: WT corn oil (Wt) (63 mice), WT tax (tax/Wt) (62 mice), PTN^−/− corn oil (PTN^−/−) (64 mice) and the PTN^−/−tax (tax/PTN^−/−) (64 mice). Following 19-21 days of blastocyst transfer, the next-generation mice were born. According to the occurrence PE of in mice during pregnancy, the mice in each group were further divided into PE and non-PE mice. Thus, there were 8 groups as follows: WT corn oil, PE/WT corn oil (PE/Wt), WT tax (tax/Wt), PE/WT tax (PE/tax/Wt), PTN^−/− corn oil (PTN^−/−), PTN^−/− corn oil (PE/PTN^−/−), the PTN^−/− tax (tax/PTN^−/−) and PE/PTN^−/− tax (PE/tax/PTN^−/−).

The blood samples were collected via venipuncture into tubes containing anticoagulants. The protein level of PTN in serum was determined on covered 96-well ELISA plates, as previously reported (32). Rabbit anti-human PTN monoclonal antibodies (1:2,000, ab14025; Abcam, Cambridge, MA, USA) were diluted in Tris-buffered saline (TBS) and incubated at 4°C overnight. Biotinylated affinity-purified anti-rabbit secondary antibody (1:50,000, ab6720; Abcam) was added into the wells and incubated at room temperature for l h. Streptavidin/alkaline phosphatase conjugate (Roche Diagnostics GmbH, Mannheim, Germany) was added and maintained at room temperature. The absorbance at 405 nm was measured on a plate reader. Recombinant human PTN (R&D Systems, Inc.) was used as a standard control. The concentration of TNF-α was also measured using an ELISA kit (ab181421; Abcam) according to the protocols. The kit mainly consists of anti-TNF-α antibodies and the regent for the ELISA assay. The absorbance of TNF-α was determined at 450 nm.

The non-invasive tail-cuff method using the CODA™ system (Kent Scientific, Torrington, CT, USA) was applied to monitor the blood pressure of mice following the steps described in a previous study (33). The measurements were all performed 5 min prior to the start of the experiment. The assessments were repeated four times for each measurement.

The urine was sampled using a bladder massage method, as described previously (34). The quantity of urine protein was determined using a Bradford assay (Bio-Rad Laboratories, Inc., Hercules, CA, USA). Standard solutions were diluted with Bradford regent, and the mixture of the two was incubated at room temperature for 5 min. The absorbance was read at 595 nm and then a standard curve was obtained. The urine protein level was assessed according to this curve.

The paraffin-embedded placental and kidney tissues from the pseudopregnant mice were dewaxed and rehydrated. For H&E staining, the sections (4-µm-thick) were stained with hematoxylin at room temperature for 10 min. Following incubation with 1% hydrochloric acid ethanol for 3 sec, the section was stained with 0.5% eosin for 30 sec. For IHC, the sections were preheated in distilled water at 37°C. The sections were first heated at 120°C in citric acid buffer and then maintained at room temperature for 20 min for antigen retrieval. Fetal calf serum (Gibco/Thermo Fisher Scientific, Waltham, MA, USA) in PBS (20%) was incubated with the sections for 10 min. The primary antibody was added onto the sections and maintained at 4°C overnight following removal of the serum. The secondary antibody (biotinylated) (Cell Signaling Technology, Inc.) was incubated for 30 min. Subsequently, peroxidase-conjugated streptavidin biotin (Cell Signaling Technology, Inc.) was added and maintained for 30 min at room temperature. Diaminobenzidine was used to identify the peroxidase activity. Finally, the sections were counterstained with Mayer’s hematoxylin (Sangon Biotech Co., Ltd., Shanghai, China). Primary antibodies used in IHC were as follows: Anti-PTN goat polyclonal (1:20, cat. no. ab223674; Abcam), anti-RPTPβ/ζ mouse monoclonal (1:200, cat. no. sc-33664; Santa Cruz Biotechnology, Inc., Dallas, TX, USA), anti-SDC1 rabbit monoclonal (1:2,000, cat. no. ab128936; Abcam) and anti-SDC3 rabbit monoclonal (1:20, ab36653, antibody). The corresponding goat anti-rabbit IgG (ab64256, 1:200) or goat anti-mouse IgG (ab64255, 1:200) secondary antibodies used were all biotinylated and purchased from Abcam.

Lysates from the placenta were denatured in boiling water for 5 min. The lysates (25 µg/lane) were mixed with loading buffer and electrophoresed on a 12% SDS PAGE gel. The proteins were then transferred onto PVDF membranes. To block the non-specific proteins, not fatty milk was incubated with the membrane for 2 h at room temperature. Anti-pleiotrophin (1:10,000, cat. no. ab79411; Abcam), anti- RPTPβ/ζ (1:800, cat. no. sc33664; Santa Cruz Biotechnology, Inc.), anti-Syndecan-1 (1:1,000, cat. no. ab181789), anti-SDC3 (1:1,000, cat. no. ab155952) (both from Abcam), anti-anaplastic lymphoma kinase (ALK; cat. no. sc-398791; Santa Cruz Biotechnology, Inc.) and anti-actin antibodies (1:1,000, cat. no. ab8226; Abcam) were used as the primary antibodies. The primary antibodies were incubated with the membranes at 4°C overnight. The appropriate HRP-conjugated IgG secondary antibody (ab6721, 1:2,000; Abcam) was incubated with the membranes at room temperature for 1 h. An ECL system (Amersham; GE Healthcare Life Sciences, Chalfont, UK) was used to detect the bands.

Data are shown as the means ± standard deviation (SD) and were analyzed using a two-tailed Student’s t-test or one-way analysis of variance followed by Tukey’s post hoc test. P<0.05 was considered to indicate a statistically significant difference. GraphPad Prism version 6.0 (GraphPad Software, Inc., La Jolla, CA, USA) was used to perform the statistical analyses.

PE in pregnancies following ART has gained increased attention (35). The clinical characteristics of the patients in the present study are summarized in Table I. It was found that the incidence of PE following vitrified-warmed embryo transfer was 13.5% (12/89). PE is considered to be systematical inflammatory responses. The concentration of TNF-α was determined by ELISA, which is closely associated with inflammation (25). Notably, as shown in Table II, increased secretion of TNF-α was observed in patients with PE than that in normal pregnant women (29±4 vs. 12±3 pg/ml, P<0.05). It was also found that the activity of PTN was reduced in patients with PE (6±13 ng/ml) when compared with the normal patients (8±17 ng/ml).

To investigate the potential influence of PTN on the incidence of PE following vitrified-warmed embryo transfer, the present study evaluated the pregnancy outcome of the conditional induced PTN-deficient mice. As shown in Table III, the knockdown of PTN did not affect the pregnancy rate, which was ~40% among the four groups. However, compared with the PTN^−/− group (P<0.05), the birth rate in the tax/PTN^−/− group decreased by almost half. In addition, the incidence of PE was significantly increased by the deficiency of PTN compared with that in the PTN^−/− group (P<0.05).

As increased hypertension and urine protein content are two principal features of PE, the present study monitored the change of blood pressure and the content of urine protein in mice prior to and following their pregnancies at certain stages. These two parameters in control mice (non-PE mice) were maintained at a normal level throughout the study. As shown in Fig. 1, the blood pressure of the PE mice remained stable prior to pregnancy and in early pregnancy (during the first 13 days of pregnancy). At day 15 of pregnancy, a significant increase in blood pressure was observed in these PE mice. In addition, the blood pressure in those PE mice that lacked PTN was higher than that in the other PE mice (PE/Wt. vs. PE/tax/PTN^−/− at 15 days: 105±5.0 vs. 112±4.8 mmHg, P<0.05). Additionally, the protein content of urine in the PE mice began to increase from day 13 of pregnancy, and was higher in the PE mice that lacked PTN than in the other mice with PE (Wt vs. tax/PTN^−/− at 13 days: 0.51±0.05 vs. 0.565±0.038 mg/day, P<0.05). From day 19 of pregnancy, the blood pressure and the urine protein content reached the highest levels in the PE mice that lacked PTN.

To determine the effect of PTN on PE following vitrified-warmed embryo transfer, the activities of PTN and TNF-α were identified in mice. As shown in Fig. 2A, the release of PTN was reduced in the PE mice as pregnancy progressed, and the release of PTN was higher in the PE mice that lacked PTN than in the other mice with PE. The secretion of TNF-α showed an increasing trend in the PE mice as pregnancy progressed. The level of PTN was low in the PE mice that lacked PTN compared with the other mice with PE, and the content of TNF-α was highest in the PE mice that lacked PTN (Fig. 2B).

PTN is involved in inflammation via its distinct receptors (36). The present study compared the expression level of PTN and its receptors in the placenta of PE and normal mice. As shown in Fig. 3, the expression levels of PTN, SDC1, SDC3 and RPTPβ/ζ were reduced in the PE mice when compared with levels in the normal mice, whereas the expression level of ALK was similar among the groups. It appears that the expression levels of SDC1 and SDC3 were not affected significantly by the knockdown of PTN, whereas the expression of RPTPβ/ζ appeared to be suppressed by the knockdown of PTN. In order to determine whether PTN and its receptors are involved in PE following vitrified-warmed embryo transfer, the expression pattern of PTN and its receptors in each group were determined by IHC (n=4). Staining for the expression of PTN was present in the cytoplasm and mesenchyme (ME) in the placenta of control mice, whereas perinuclear and ME staining was weaker in the placenta of PE mice (Fig. 4). The PTN staining faded out in mice that lacked PTN, although PTN staining remained present in the tax/ PTN^−/− mice as the efficiency of the conditional knockout was not 100% effective (negative staining for the isotype control of PTN; data not shown). The expression of PTN was almost absent in the tax/PTN^−/−PE mice (Fig. 4D). In the placental tissues of the control mice, strong fetal macrophage (FM) staining for SDC1 was also observed. Fainter staining was observed in the ME and fetal vessels (FV). In the PE mice, not all FMs were stained and other staining had almost disappeared (Fig. 5). The expression ofSDC1 appeared to be reduced in the PTN-knockout mice (Fig. 5D). Strong diffuse cytoplasmic staining for RPTPβ/ζ was also noted. Intensive staining in the ME, villous cytotrophoblasts, syncytiotrophoblasts and FMs were observed in the control mice. However, this staining was weaker or absent in the placenta of the PE mice (Fig. 6).

Renal damage is often connected to PE (37). Therefore, the present study examined kidney injury in the mice. As shown in Fig. 7, the glomeruli, renal capsule and tubular structures were clear in the kidneys of the control mice. However, the Bowman’s capsule and the opening of capillary loops were reduced, and inflammatory cell infiltration increased in the PE mice. In addition, kidney injury was more severe in the PTN-knockout PE mice.