The experimental protocol was approved by the Institutional Review Board of Shengjing Hospital of China Medical University (Shenyang, China). All the experiments were conducted according to the institutional guidelines for the humane treatment of laboratory animals. In the present study, a high-fat diet combined with intraperitoneal injection of streptozotocin (STZ; an antibiotic that produces pancreatic islet β-cell destruction) was employed to generate a rat model of GDM (26-28). A total of 70 (10 male and 60 female) Sprague-Dawley rats aged 9-10 weeks were purchased from Changsheng Biotechnology Co., Ltd. The mean weight of the rats was 301±52 g. All the animals were specific-pathogen free, housed at 21-22˚C, under a relative humidity of 60-70% and a 12-h light/dark cycle. The female rats were randomly divided into two groups (30 rats per group), namely a high-fat diet group and a normal diet group. The high-fat diet group was fed a high-fat diet (Huafukang Medical Technology Co., Ltd.) with a daily fat content of 60%, whereas the normal diet group was fed a normal diet. All rats had free access to food and autoclaved water.

After one month, female and male rats were placed in cages at a 2:1 ratio and the vaginal plug or vaginal secretion smear of the female rats was examined the next morning. The day on which sperm or vaginal plugs were observed in the rats was recorded as pregnancy day 0. Following the detection of pregnancy in the high-fat diet group, rats were starved for 12 h with provision of water ad libitum and then administered with an intraperitoneal injection of STZ (cat. no. S0130; MilliporeSigma) at a dosage of 30 mg/kg body weight (dissolved in 0.1 mmol/l sodium citrate, ready to use). The rats were then fed the high-fat diet. One week later, 100 µl blood from each rat was collected through the tail vein to evaluate fasting blood glucose and non-fasting blood glucose; measurements were conducted three times separately. Rats with ≥120 mg/dl fasting blood glucose and at the same time, ≥300 mg/dl non-fasting blood glucose were diagnosed with GDM (29-32); measurements were conducted three times separately. The normal diet group continued to receive the normal diet.

After confirming pregnancy, the blood pressure of each rat was monitored every week using the tail-cuff method (using a Softron BP-98A and a Softron TMC-203). Three days before the formal measurement, the blood pressure was measured at a fixed time every day for adaptation to measurement conditions. The first three measurements were not recorded as a formal measurement and the measurement commenced only after the rats were stabilized. Blood pressure in each rat was measured three times and the mean value was recorded (14). Tail pressure was monitored every week until pregnancy day 20, as the average gestational period of rats is 21 days. Currently, no uniform standards exist for normal blood pressure and hypertension in pregnant rats. In the present study, hypertension was defined as systolic blood pressure elevated by >20 mmHg compared with the initial measurement (33-35). Based on the blood pressure values, rats with GDM were divided into a GDM group (GDM) and a GDM with hypertension group (GDM + HT). The blood pressure measurements were performed repeatedly to reduce error. The normal diet group served as the control group (CONT; Fig. 1).

Rats in all three groups were euthanized by exsanguination via abdominal aorta after anaesthetising by isoflurane inhalation (2.5-3.5% for the induction phase and 2.5-3.0% for the maintenance phase) on pregnancy day 20. Successful euthanasia was confirmed via respiratory and cardiac arrest. Following euthanasia, the abdominal aorta was collected immediately.

During sample collection, care was taken to avoid damage to the vascular endothelium. The tissue samples from the three groups of rats were fixed in 10% formaldehyde and ethanol at 25˚C for 24 h. Following dehydration and paraffin embedding, 4-µm serial sections were obtained. After heating of paraffin-embedded sections in a 60˚C incubator for 0.5-1 h, sections were deparaffinized with xylene at 25˚C, rehydrated in a concentration descending ethanol series and immersed in endogenous peroxidase blocking buffer for 10 min. Next, antigen retrieval was performed using high-pressure antigen retrieval. Diluted primary antibodies, BMP-4 (cat. no. ab39973; 1:100), COX-2 (cat. no. ab15191; 1:2,000), VCAM-1 (cat. no. ab134047; 1:500) and NOX-1 (cat. no. ab131088; 1:1,000; all obtained from Abcam) were added and slides were incubated overnight at 4˚C. PBS treatment served as a negative control. Slides were re-heated at 37˚C for 20 min and rinsed three times with PBS solution. The reaction enhancer was added and slides were incubated at 25˚C for 20 min and rinsed three times with PBS solution. The horseradish enzyme labelled goat anti-rabbit IgG (cat. no. ZB-2301; 1:5,000; OriGene Technologies, Inc.) was added and slides were incubated at 25˚C for 20 min. Slides were then rinsed three times with PBS solution. Subsequently, 3,3'-diaminobenzidine was used for colour development. Slides were rinsed with running water and haematoxylin was used for counterstaining of nuclei, while hydrochloric acid and ethanol were used for differentiation. Finally, slides were rinsed with running water, dehydrated and dried using an ethanol gradient, cleared with xylene and mounted with neutral gum.

The BMP-4, NOX-1, COX-2 and VCAM-1 proteins are located primarily in the cytoplasm of vascular endothelial cells and positive expression is indicated by brown or tan colouration in immunohistochemistry. Five high-magnification (x400) fields of view were randomly selected for each slide, and the staining intensity and number of positive cells were observed. A score was assigned based on the staining intensity as follows: 0, No staining; 1, light yellow; 2, brown; 3, tan. Another score was assigned based on the percentage of positive cells as follows: 0, All cells negative; 1, 1-10% positive cells; 2, 11-50% positive cells; 3, 51-75% positive cells; and 4, >76% positive cells. A positive immune reaction was defined as a product of the staining intensity score and the percentage of positive cells >3.

Rat aortas were harvested after the indicated treatment. Protein was extracted using 300 µl lysis buffer (cat. no. WLA019a; Wanleibio Co., Ltd.) supplemented with protease and phosphatase inhibitors. BCA method was used to determine the concentration of the protein and 50 µg protein was loaded per lane. Western blotting was performed according to standard methods: Protein separation by 10% gel electrophoresis, transferring the protein from the gel onto polyvinylidene fluoride membranes (Bio-Rad Laboratories, Inc.), the membranes were blocked using 3% bovine serum albumin (cat. no. B2064; Sigma-Aldrich; Merck KGaA) at 25˚C for 1 h, and then were probed with BMP-4, COX-2, VCAM-1 and NOX-1 antibodies (dilution of all the antibodies was 1:1,000), β-actin antibody (cat. no. sc-1616; Santa Cruz Biotechnology, Inc.; 1:1,000) was used for the reference protein, the membranes were incubated overnight at 4˚C, then the goat anti-rabbit IgG (cat. no. sc-2004; Santa Cruz Biotechnology, Inc.; 1:1,000) was added in the membranes and then the membranes were incubated at 25˚C for 1 h, then an enhanced chemiluminescence method using chemiluminescent substrate (cat. no. 34577; Thermo Fisher Scientific, Inc.) was performed. The molecular weight of each protein shown on the immunoblot was estimated based on the Rainbow 245 Marker for Western Blotting Protein Standard (Beijing Solarbio Science & Technology Co., Ltd.) on a 10% SDS-PAGE gel. Quantification was performed based on the density of bands obtained using ImageJ v1.8.0 Software (National Institutes of Health). β-actin served as the loading control.

Following fertilization, female rats that were fed with a normal diet were randomly divided into four groups (n=4 rats per group) designated as A, B, C and D. Blood pressure was measured using the tail-cuff method on the day pregnancy was discovered and recorded as the initial blood pressure. In order to evaluate whether BMP-4 affected the blood pressure of pregnant rats, osmotic minipumps (Alzet^®Osmotic Pumps; Durect Corp.) were implanted using 1% sodium pentobarbital as the anaesthetic (3 ml/kg via intraperitoneal injection) on the third day of pregnancy. The solution was composed of an infusion vehicle, recombinant human BMP-4 (cat. no. 314-BP-050; R&D Systems, Inc.) or recombinant human noggin (cat. no. 6057-NG-100; R&D Systems, Inc.) supplemented with 0.1% bovine serum albumin (cat. no. A8020, Solarbio) in 4 mmol/l HCl. Four groups were evaluated as follows: Group A was infused with 0.45 mg/kg BMP-4 + vehicle, group B was infused with 0.45 mg/kg noggin + vehicle, group C was infused with 0.45 mg/kg BMP-4 + vehicle and 0.45 mg/kg noggin + vehicle, two osmotic minipumps were used and implanted at different sites for infusion to avoid mixing, and group D was infused with vehicle alone and served as a control (14). These four groups of rats were administered with infusions continuously for 2 weeks. Blood pressure was monitored and recorded again on pregnancy day 18 and the changes in blood pressure among the four groups were compared.

Rats in all four groups were euthanised by exsanguination via abdominal aorta after anaesthetising by 2.5-3.0% isoflurane inhalation on pregnancy day 20. Following euthanasia, the thoracic aorta was collected immediately for vascular reactivity testing. During sample collection, care was taken to avoid damage to the vascular endothelium. After collection, the aorta was immediately placed in a petri dish containing Krebs-Henseleit (KH) solution (cat. no. PB180348; Procell Life Science & Technology Co., Ltd.) at 37˚C. A mixture of 95% oxygen + 5% carbon dioxide gas was continuously injected into the KH solution. The fatty tissue on the surface of the blood vessels was removed and blood vessels were cut into rings (length, 4 mm). Two L-shaped stainless-steel hooks were passed through the lumen, with one end fixed and the other end connected to the tension transducer of a physiological recorder (PL3508 PowerLab 8/35; ADInstruments, Ltd.). Blood vessels were immersed in the KH solution and the physiological recorder was adjusted to apply a baseline tension of 1 g to the blood vessels to simulate physiological tension. The tension was equilibrated for 60 min, during which the KH solution was replaced every 15 min. After stabilisation of the tension, 10^-5 mol/l phenylephrine was added to the container, and 10^-9-10^-6 mol/l acetylcholine was added after vasoconstriction was equilibrated. The time interval between additions of different concentrations of acetylcholine was 2 min. The rings were continuously aerated while phenylephrine and acetylcholine were added. The relaxation of the thoracic aorta vascular ring of each group of rats was recorded.

SPSS 23.0 software (IBM Corp.) was used for statistical analysis and data are expressed as means ± standard error. Student's t-test was used for comparisons between two groups and one-way analysis of variance was used for comparisons between multiple groups. Tukey's post hoc test was performed for pairwise comparisons between groups. The χ² test or Fisher's exact probability test was performed for comparisons of countable data. P<0.05 was considered to indicate a statistically significant difference.

As shown in Fig. 2, the positive expression rates of BMP-4 in the GDM, GDM + HT and control groups were 100% (11/11), 100% (4/4) and 72.7% (8/11), respectively. The positive expression rates of NOX-1 in the GDM, GDM + HT and control groups were 100% (11/11), 100% (4/4) and 81.8% (9/11), respectively. The positive expression rates of COX-2 in the GDM, GDM + HT and control groups were 100% (11/11), 100% (4/4) and 72.7% (8/11), respectively. The positive expression rates of VCAM-1 in the GDM, GDM + HT and control groups were 100% (11/11), 100% (4/4) and 10% (1/10), respectively. The expression levels of all four proteins in the control group were lower than those in the GDM and GDM + HT groups. The positive expression rate of VCAM-1 in the control group was significantly lower than that in the GDM and GDM + HT groups; the differences were statistically significant (χ²=17.325, P<0.05 and χ²=10.080, P<0.05, respectively).

The expression level of the BMP-4 protein in the control, GDM and GDM + HT groups exhibited a sequentially increasing tendency; however, the difference observed among the groups was not statistically significant (F=2.651, P=0.13). Similarly, the difference in the expression levels of the NOX-1 protein among the three groups was not statistically significant (F=2.122, P=0.18). By contrast, the expression levels of COX-2 protein in the three groups exhibited a sequential increase and the difference was observed to be statistically significant (F=8.560, P<0.05). The difference in the expression levels of the COX-2 protein between the control (1.000±0.255) and the GDM (1.434±0.344) groups was not identified to be statistically significant (P=0.78), but the expression levels in the control and GDM groups were significantly lower than those in the GDM+ HT group (3.358±1.286, P<0.05 and P<0.05, respectively). The expression levels of the VCAM-1 protein in the three groups also exhibited a sequential increase, with a statistically significant difference (F=31.732, P≤0.001). The difference in the expression levels of the VCAM-1 protein between the control (1.000±0.297) and GDM (1.525±0.347) groups was not statistically significant (P=0.11), but the expression levels in the control and GDM groups were significantly lower than those in the GDM + HT group (2.698±0.223, P≤0.001 and P≤0.001, respectively; Fig. 3).

Fig. 4 demonstrates that infusion of BMP-4 significantly increased the blood pressure of rats by 30 mmHg above the control values (from 82.3±5.0 to 112.3±6.3 mmHg; P<0.05).

The specificity of the hypertensive effect of BMP-4 was evaluated by infusing noggin, a BMP-4 antagonist, either alone or in combination with BMP-4. As shown in Fig. 4, noggin co-delivery prevented the BMP-4-induced hypertension in pregnant SD rats.

To evaluate the effect of BMP-4 on vascular endothelial function in pregnant rats, the diastolic activity of the thoracic aorta of rats in each group were assessed. The results demonstrated that the maximum relaxation in the control group after adding acetylcholine was 88.90±1% of the maximum contractile tension, whereas the maximum relaxation in the BMP-4-infused group after adding acetylcholine was 68.96±0.8% of the previous maximum contractile tension, which was significantly smaller than that in the control group (P<0.05). Accordingly, it was observed that the blood vessels in the noggin-infused group were nearly the same as those in the control group (Fig. 5 and Table I).