Effect of Metformin on a Preeclampsia-Like Mouse Model Induced by High-Fat Diet.

BACKGROUND: Metformin has been reported to decrease insulin resistance and is associated with a lower risk of pregnancy-induced hypertension and preeclampsia. It is widely accepted that the placenta plays a crucial role in the development of preeclampsia. Our aim is to explore the effect of metformin on preeclampsia. STUDY
DESIGN: We examined control diet-fed (isocaloric diet) pregnant mice (CTRL group), pregnant mice fed a high-fat diet (HF group), and high-fat-diet-fed pregnant mice treated with metformin (HF-M group). The HF mice were fed a high-fat diet six weeks before pregnancy to establish a preeclampsia-like model; then, the group was randomly divided into a HF group and a HF-M group after pregnancy. Blood pressure, urine protein, pregnancy outcomes, protein expression, and histopathological changes in the placentas of all groups were examined and statistically analysed.
RESULTS: We observed that metformin significantly improved high blood pressure, proteinuria, and foetal and placental weights in the HF-M group compared with the HF group. Metformin significantly improved placental labyrinth and foetal vascular development in preeclampsia. In addition, metformin effectively increased matrix metalloproteinase-2 (MMP-2) and vascular endothelial growth factor (VEGF) levels in the placenta.
CONCLUSIONS: Our results suggest that metformin can improve preeclamptic symptoms and pregnancy outcomes.

1. Introduction

Preeclampsia is characterized by the onset of high blood pressure and proteinuria and results in substantial maternal and neonatal morbidity and mortality [1-3]. Type 1 and type 2 diabetes (T2DM), gestational diabetes (GDM), and polycystic ovarian syndrome (PCOS) are also risk factors for preeclampsia [4]. Specifically, women with T2DM have high rates of maternal morbidity, including gestational hypertension and preeclampsia (17–19%) [5-8]. Moreover, some reports have demonstrated that an increased body mass index (BMI) increases the risk of preeclampsia and that obesity is an important risk factor for preeclampsia [9, 10].

Metformin (1,1-dimethylbuguanide hydrochloride), an oral biguanide insulin sensitizer, is a low-cost, low-risk, effective and approved oral hypoglycaemic agent for T2DM and is an important treatment option for patients with GDM [11, 12]. An increasing number of patients with PCOS or diabetes have used metformin during pregnancy, and researchers have found that metformin is associated with a lower risk of pregnancy-induced hypertension [11, 13]. Importantly, metformin is associated with a lower risk of neonatal hypoglycaemia and neonatal intensive care admission [14]. The available data also suggest that metformin exposure during the first trimester is not associated with major congenital malformations but reduces the risks of early pregnancy loss, preeclampsia, preterm delivery, and GDM, prevents or attenuates antipsychotic-associated weight gain, and improves conception chances and pregnancy outcomes in the presence of PCOS [14-20].

It is widely accepted that preeclampsia is associated with the two-stage model of the placenta. Extravillous cytotrophoblast invasion (EVT) of the uterine spiral arteries is limited to the superficial decidua and does not reach the myometrium, which is believed to lead to shallow implantation and hypoperfusion of the placenta during the first stage of pregnancy. In the second and third trimesters, hypoperfusion of the placenta results in placental hypoxia and ischaemia, and the maternal-foetal interface is subjected to relatively strong oxidative stress. Subsequently, cytokines and cell debris are secreted into the circulation of the maternal blood system by the placenta, eventually leading to maternal preeclamptic symptoms and injuries to other important organs [21-25].

Although researchers have investigated the use of metformin during pregnancy, the precise role of metformin in the placenta of preeclampsia remains unknown. Moreover, it has been demonstrated that insulin resistance together with adipocyte dysfunction might be involved in the pathophysiology of preeclampsia in obese women [26-28]. A literature search revealed a study in which high-fat-diet-fed pregnant mice exhibited significantly worse glucose tolerance and insulin sensitivity than control diet-fed pregnant mice, as shown by their elevated blood glucose levels, fasting serum insulin levels, and HOMA-IR, which were also increased in the HF group compared with the CTRL group; moreover, researchers have successfully fed mice a high-fat diet to induce preeclampsia [29-31].

Increasing attention has been paid to the application of metformin during pregnancy, and many studies have been carried out; however, to our knowledge, there is no study on the effect of metformin on the placenta during preeclampsia. The mechanism of metformin in reducing the incidence of preeclampsia is still unclear. Therefore, we employed a high-fat-diet-induced mouse model of preeclampsia and administered metformin to investigate the pregnancy outcomes of preeclampsia and the histological changes in the placenta during preeclampsia. We also examined the protein levels of matrix metalloproteinase (MMP)-2 and vascular endothelial growth factor (VEGF) to confirm the possible mechanism of metformin in preeclampsia. We further explored the effect of metformin on preeclampsia and provide evidence for the use of metformin during pregnancy.

2. Materials and Methods

2.1. Establishment of Animal Model

This study was approved by the Animal Ethics Committee of the Capital Medical University Beijing Chao-Yang Hospital. Animals were given humane care in accordance with the guide for the care and use of experimental animals. Thirty-eight female, 4-week-old CD-1 mice were obtained from Weitong Lihua Experimental Animal Technology Co., Ltd. (Beijing, China) and randomly divided into the CTRL group (n = 13), which was fed an isocaloric stock diet (energy ratio: 12% fat, 28% protein and 60% carbohydrates), and the HF group (n = 25), which was fed a high-fat diet (energy ratio: 62% fat, 18% protein, 20% carbohydrates) purchased from Research Diets, Inc. (New Brunswick, USA). All animals were housed in groups of 4–6 mice per cage, under an ambient temperature of 22–26°C, a relative humidity of 50–60%, and a 12 h-12 h light-dark cycle and fed their respective diets (ad libitum). After 6 weeks, the mice were weighed and mated. Females were checked daily for postcopulatory plugs, and the presence of a plug represented day 0.5 of pregnancy. The twenty-five HF mice were randomly divided into 2 groups, namely, the HF group and the HF-M group, starting at day 0.5 of pregnancy. Metformin was administered to the HF-M group at a daily dose of 20 mg/kg. Blood pressure was measured via the tail-cuff method using a softron BP-98A tail-cuff haemodynamometer (Softron, Tokyo, Japan) after the behaviour and heart rate of the mice had stabilized. The blood pressure is reported as the mean of at least three measurements recorded during the same session that exhibited <5% variation. Most blood pressure values were within the required range once the mice had stabilized. Urine samples were collected by the reflex voiding method at day 18.5 of pregnancy; the urinary protein level was measured using a mouse albumin ELISA quantitation set (Bethyl Corp USA); urinary creatinine was detected using mouse urine creatinine detection kits (Cayman Corp USA), according to the manufacturer's instructions. All mice were sacrificed by vertebral dislocation at day 18.5 of pregnancy. The foetal and placental weights were measured, and the placentas were removed immediately. Part of the placenta was immediately frozen and stored at −80°C, and the other part of the placenta was fixed in 10% formaldehyde.

2.2. Histological Analyses

The placental specimens were dehydrated in a graded alcohol series and embedded in paraffin. Sections were deparaffinized and stained with haematoxylin-eosin (HE) for histological examination. Twenty microscopic fields (400x) were examined using a microscope (Olympus Corp, Japan).

2.3. Immunohistochemical Analysis

Paraffin sections of the placenta were deparaffinized, rehydrated, and subjected to antigen retrieval in 0.01 M sodium citrate buffer (pH 6.0) in a microwave oven. Then, the sections were treated with 3% hydrogen peroxide to block endogenous peroxidase activity. The placental sections were subsequently incubated with primary polyclonal antibodies against laminin (Santa Cruz Biotechnology, Santa Cruz, CA, USA), diluted 1 : 100 and 1 : 50. After washing with tris-buffered saline (TBS), specific secondary antibodies were applied, and then, the sections were incubated with an avidin-biotin-peroxidase complex according to the manufacturer's instructions (ABC-peroxidase kit, Vector Labs, Burlingame, CA, USA). Finally, the sections were incubated in a solution of 0.05% 3–3′-diaminobenzidine tetrahydrochloride (Sigma Aldrich, St. Louis, MO, USA) and 0.33% hydrogen peroxide. All sections were counterstained with Harris haematoxylin, dehydrated in a graded alcohol series and in xylene, and coverslipped.

2.4. Western Blotting

The proteins of the placentas were extracted, and the protein concentration was determined with the bicinchoninic acid protein quantification method. Then, 30 μg of total protein was separated by SDS-PAGE and electrotransferred onto nitrocellulose membranes (Millipore, Bedford, MA, USA) using a semidry western blot transfer system (Bio-Rad Laboratories, Inc., Hercules, CA, USA). The membranes were incubated with primary rabbit anti-MMP-2 monoclonal antibody (1 : 1,000 dilution), rabbit anti-VEGF monoclonal antibody (1 : 1,000 dilution) (Abcam, Cambridge, MA, USA), and anti-GAPDH (loading control, 1 : 3,000 dilution) at 4°C overnight. After washing with PBST three times, the membranes were incubated with secondary HRP-conjugated goat anti-rabbit antibody (1 : 10,000 dilution; Cell Signalling Technology, Inc.) for 2 h at room temperature. Finally, the signals were developed with a chemiluminescent ECL reagent (Millipore), and the membrane was exposed to X-ray film. The strips were analysed by scanning using Image Lab software. The ratio of the protein band density to the reference GAPDH band density was determined as the relative expression level of the target protein.

2.5. Statistical Analysis

Statistical analyses were performed by two-way ANOVA for comparisons among the 3 pregnant mouse groups, and unpaired t tests were used for comparisons between two groups. GraphPad Prism software, version 4.0 (GraphPad Software, Inc., San Diego, CA, USA), was used for data analysis and plotting. Data are presented as the mean ± SD. Values of P < 0.05 were considered to indicate significant differences.

3. Results

3.1. Effects of Metformin on Maternal Symptoms of Preeclampsia

The blood pressure and urinary protein levels in the HF group were significantly elevated compared to those in the CTRL and HF-M groups. The blood pressure in the HF-M group began to decrease on the 12th day, compared with that in the HF group; there was a significant difference on days 12, 14, 16, and 18. Metformin improved blood pressure and urinary protein levels in the HF-M group (both P < 0.05) (Figure 1).

Figure 1

Effects of treatment with metformin on maternal signs of preeclampsia. (a) Systolic blood pressure. (b) Diastolic blood pressure. (c) Mean arterial blood pressure. (d) Urinary protein. The results are presented in mean ± SD. P < 0.05.

3.2. Subject Characteristics

The characteristics of the mice are summarized in Table 1.

Table 1

Characteristics of the mice.

Characteristics	CTRL (n = 13)	HF (n = 13)	HF-M (n = 12)
Body weight before pregnancy (g)	29.38 ± 1.44	32.92 ± 2.50	32.50 ± 2.24
Body weight before delivery (g)	56.08 ± 4.29	56.30 ± 4.33	55.08 ± 3.53
Litter size of each group	174	175	164
Foetal weight (g)	1.06 ± 0.19	0.95 ± 0.17	1.02 ± 0.16
Placental weight (g)	0.11 ± 0.02	0.09 ± 0.01	0.10 ± 0.02

Data indicate the mean ± SD.

3.3. Effects of Metformin on Maternal, Foetal, and Placental Weights

A significant difference in body weight before pregnancy was observed between the model group and the CTRL group mice (P < 0.05). No significant differences in body weight at day 18.5 of pregnancy were found among the three groups (P > 0.05). The effect of metformin on the foetoplacental unit was also examined. The foetal and placental weights of the HF group were significantly lower than those of the CTRL and HF-M groups (P < 0.05), but no significant difference in foetal weight was observed between the CTRL and HF-M groups (P > 0.05). Metformin treatment enhanced the placental and foetal weights of the HF group (P < 0.05). No significant difference in the average litter size was observed among the three groups (12.9 ± 0.4) (see Figure 2).

Figure 2

The weights of (a) the pregnant mice, (b) the placenta, and (c) the foetus at day 18.5 of pregnancy. Data are expressed as the mean ± SD. P < 0.05 between two groups.

3.4. Analysis of the Placental Structure of the Pregnant Mice

We analysed the tissue structure of the placentas. HE staining of placental tissue sections showed that the overall cross-sectional area of the placenta was obviously decreased in the HF group but was improved after metformin treatment, which is consistent with the decrease in placental weight (Figure 3(a)). We also analysed the labyrinth and sponge of the placenta. The labyrinth of the placenta in the HF group was significantly smaller (Figure 3(b)). There was no significant difference in the sponge of the placenta (Figure 3(c)). The ratio of the labyrinth to the sponge of the placenta in the HF group was significantly smaller than that in the CTRL group (Figure 3(d)). Furthermore, laminin labelling of foetal blood vessels in the foetal vascular branches of the HF group was significantly affected (Figure 3(e)). The foetal blood vessels and the maternal blood sinusoids in the labyrinth layer were marked (Figure 3(e), red indicates foetal blood vessels, and blue indicates the maternal sinusoids). Statistical analysis of the foetal blood vessel area and maternal blood sinuses showed that the maternal blood sinus area was not affected in HF mice (Figure 3(f)) but that the foetal vascular area in the labyrinth layer of the preeclamptic group was decreased to 15% of that in the CTRL group (Figure 3(g)).

Figure 3

Analysis of placental structure. (a) Placental HE staining results of the three groups. The magnification is 20x in the upper panels. (b–d) Bar charts show the labyrinth and sponge of the placenta. (e) An immunohistochemical analysis of laminin in the placenta. The magnification is 400x in the panels. Pseudocolour is used to indicate foetal blood vessels (red) and the maternal blood sinusoid (blue) in the placental labyrinth. (f–g) Bar graphs show the quantitative results of the red and blue areas of all placentas in each group, indicating the area of the maternal blood sinusoid (MBS) and foetal blood vessels (FBV). The statistical analysis was based on the results of all placentas in each group. Data are expressed as the mean ± SD. P < 0.05 between two groups.

3.5. Effect of Metformin on Placental Proteins

We further analysed the expression of proteins in the placenta by western blotting. Our results showed that the levels of VEGF and MMP-2 were significantly decreased in the HF group (P < 0.05). However, after metformin treatment, the levels of VEGF and MMP-2 in the placentas of preeclamptic mice were significantly increased (P < 0.05) (see Figure 4).

Figure 4

The protein expression of the placenta by western blotting. (a) Western blotting shows the relative expression of VEGF and MMP-2 in the placentas. (b-c) Bar graphs show the relative protein levels of VEGF and MMP-2 based on the statistical analysis of the western blotting results. Data are expressed as the mean ± SD. P < 0.05 between two groups.

4. Discussion

Metformin can increase the insulin sensitivity. Studies have indicated that insulin resistance is one of the causes of preeclampsia [32]. Among studies in the literature, there is no study on metformin in preeclampsia patients. In one in vitro study on the prevention and treatment of preeclampsia with metformin, trophoblasts and omental vessels, isolated from the placentas of 23 pregnant women with severe preeclampsia and 25 women with normal pregnancies, were co-cultured with metformin, and researchers found that metformin could significantly improve the dysfunction in placental endothelial cells in preeclampsia, dilate and ameliorate injured vessels, and promote angiogenesis [17, 32].

Our study successfully established the signs of preeclampsia, including hypertension and proteinuria, in pregnant rats in the HF group. This is similar to the methods reported in the literature [29-31]. We also observed that the foetal weights of the preeclamptic mice were significantly lower than those of the CTRL mice, suggesting that foetal development in the preeclamptic pregnant mice was obviously delayed. This finding is completely consistent with the symptoms that we observed in patients with preeclampsia. We also found that the embryonic weight in the HF-M group increased significantly compared with that in the HF group. The weight of a new-born is an important biological parameter and reflects the growth and nutritional status of the foetus in utero. This suggests that metformin can significantly improve foetal development in preeclamptic pregnant mice.

It is well known that the majority of foetuses from preeclamptic patients show intrauterine growth restriction, and preeclampsia originates from abnormal placental development. Now, people are convinced that the placenta plays a crucial, causative role in the potential mechanism of preeclampsia. Therefore, it is now clear that preeclampsia is a placental disease. We also found that the placental weight of the HF group was significantly lower than that of the CTRL group, and the placentas of preeclamptic pregnant mice appeared retarded. The placental weight of the HF-M group dramatically increased. Nutrients are exchanged in the labyrinth of the mouse placenta, and the maternal-foetal barrier includes two layers of syncytiotrophoblasts and foetal vascular endothelial cells.

Therefore, we analysed the placental structures of the three groups. Our results show that the area of the labyrinth and the area of the foetal blood vessels in the placenta of the HF group were significantly reduced compared with those of the normal CTRL group, but those of the HF-M group increased significantly compared with those in the HF group. These results suggest that metformin can improve the development of the labyrinth, the foetal blood vessels, and the shallow implantation of the placentas of preeclamptic mice.

It is important that further efforts be made to study the possible mechanism of metformin in preeclampsia. Therefore, we examined the protein levels of the cytokines VEGF and MMP-2, which have been shown to be closely related to the development of the placenta in preeclampsia, in the placental tissues of mice by western blotting [33-36]. MMP-2, also known as gelatinase A, is the most common MMP and is thought to play important roles in trophoblast invasion.

VEGF, which is indispensable for normal pregnancy, is the most important growth factor in promoting angiogenesis. First, VEGF participates in angiogenesis and placental vascular recasting. Second, VEGF plays an important role in regulating the infiltration, proliferation, and differentiation of trophoblast cells [35, 36].

We observed that the expression of VEGF and MMP-2 in the HF group was significantly decreased compared to that in the CTRL group but that the expression levels of VEGF and MMP-2 in the HF-M group were significantly enhanced compared to those in the HF group. Therefore, we speculated that metformin might promote the expression of the VEGF and MMP-2 proteins, further improve shallow placental implantation in preeclamptic mice, and eventually ameliorate the maternal and foetal outcomes of preeclamptic mice.

Our research suggests that metformin plays an important role in the regulation of placental implantation in preeclampsia. These results will provide important information on the prediction and treatment of preeclampsia.

5. Conclusion

Metformin can significantly improve placental development in preeclamptic mice, thereby improving the related symptoms of preeclampsia and maternal and foetal outcomes, including hypertension, proteinuria, and foetal birth weight.